#install.packages("fitdistrplus") #install.packages("logspline") #install.packages("effsize") #install.packages("betareg") #install.packages("ggplot2") #install.packages("PairedData") #install.packages("tidyr") #install.packages("tidyverse") #Please change this path to your working directory setwd("/Users/Andreina/Desktop/Genie_new/manuscript_all") library(fitdistrplus) library(logspline) library(effsize) library(betareg) library(ggplot2) library(PairedData) library(RColorBrewer) library(ggpubr) library(dplyr) options(tidyverse.quiet = TRUE) library(tidyverse) library(glue) display.brewer.all(n=NULL, type="all", select=NULL, exact.n=TRUE, colorblindFriendly=TRUE) #Demographic analyses demographics_2016 <- read.table("Sfile4_BIO345_DEMOGRAPHICS_2016.txt", header = TRUE, sep = "\t") head(demographics_2016) demographics_2017_GENIE <- read.table("Sfile5_BIO345_DEMOGRAPHICS_2017_GENIE.txt", header = TRUE, sep = "\t") head(demographics_2017_GENIE) demographics_2017_NOGENIE <- read.table("Sfile6_BIO345_DEMOGRAPHICS_2017_NOGENIE.txt", header = TRUE, sep = "\t") head(demographics_2017_NOGENIE) ############################################################################################## #Counting the number of invididuals per demographic group per class year <- rep("2016", 406) demographics_2016 <- cbind(demographics_2016, year) demo2016 <- select(demographics_2016, time, Gender, First_Gen, Ethnicity, year) year <- rep("2017G", 280) demographics_2017_GENIE <- cbind(demographics_2017_GENIE, year) demo2017G <- select(demographics_2017_GENIE, time, Gender, First_Gen, Ethnicity, year) year <- rep("2017NG", 232) demographics_2017_NOGENIE <- cbind(demographics_2017_NOGENIE, year) demo2017NG <- select(demographics_2017_NOGENIE, time, Gender, First_Gen, Ethnicity, year) total <- rbind(demo2016, demo2017G, demo2017NG) Ethinicy_contingency_table <- table(total$Ethnicity, total$year) Gender_contingency_table <- table(total$Gender, total$year) First_Gen_contingency_table <- table(total$First_Gen, total$year) #Writing output to text sink('T1_demographic_output.txt', append=TRUE) Ethinicy_contingency_table Gender_contingency_table First_Gen_contingency_table sink() ############################################################################################ #Are scores demographic dependent on either 2016 or 2017 classes? demographics_2016_glm <- read.table("Sfile7_BIO345_DEMOGRAPHICS_2016.txt", header = TRUE, sep = "\t") head(demographics_2016_glm) demographics_2017_glm <- read.table("Sfile8_BIO345_DEMOGRAPHICS_2017.txt", header = TRUE, sep = "\t") head(demographics_2017_glm) #Pre-recitation scores 2016 glm_demo_2016_pre <- lm(score_pre ~ First_Gen * Ethnicity * Gender, data=demographics_2016_glm) anova(glm_demo_2016_pre) #Post-recitation scores 2016 glm_demo_2016_post <- lm(score_post ~ First_Gen * Ethnicity * Gender, data=demographics_2016_glm) anova(glm_demo_2016_post) #Pre-recitation scores 2017 glm_demo_2017_pre <- lm(score_pre ~ First_Gen * Ethnicity * Gender * Genie_used, data=demographics_2017_glm) anova(glm_demo_2017_pre) #Post-recitation scores 2017 glm_demo_2017_post <- lm(score_post ~ First_Gen * Ethnicity * Gender * Genie_used, data=demographics_2017_glm) anova(glm_demo_2017_post) #Writing output to text sink('T2_anova_output.txt', append=TRUE) cat("ANOVA on pre-recitation scores 2016\n") anova(glm_demo_2016_pre) cat("\n\n") cat("=============================\n") cat("ANOVA on post-recitation scores 2016\n") anova(glm_demo_2016_post) cat("\n\n") cat("=============================\n") cat("ANOVA on pre-recitation scores 2017\n") anova(glm_demo_2017_pre) cat("\n\n") cat("=============================\n") cat("ANOVA on post-recitation scores 2017\n") anova(glm_demo_2017_post) sink() ############################################################################################ #Input the results of both years and treatments combined_years <- read.table("Sfile9_COMBINED_TABLE_FOR_DF.txt", header = TRUE, sep = "\t") head(combined_years) #Splits the data frame into years and treatments split_years <- split(combined_years, combined_years$year) df_2016 <- split_years[[1]] head(df_2016) df_2017_G <- split_years[[2]] head(df_2017_G) df_2017_NG <- split_years[[3]] head(df_2017_NG) #Fitting the data to a distribution descdist(df_2016$score_pre, discrete = F) #beta distributionn descdist(df_2016$score_post, discrete = F) #beta distribution fit.beta_pre_2016 <- fitdist(df_2016$score_pre, "beta", method = "mme") fit.beta_post_2016 <- fitdist(df_2016$score_post, "beta", method = "mme") plot(fit.beta_pre_2016) plot(fit.beta_post_2016) descdist(df_2017_G$score_pre, discrete = F) #beta distributionn descdist(df_2017_G$score_post, discrete = F) #beta distribution fit.beta_pre_2017_G <- fitdist(df_2017_G$score_pre, "beta", method = "mme") fit.beta_post_2017_G <- fitdist(df_2017_G$score_post, "beta", method = "mme") plot(fit.beta_pre_2017_G) plot(fit.beta_post_2017_G) descdist(df_2017_NG$score_pre, discrete = F) #beta distributionn descdist(df_2017_NG$score_post, discrete = F) #beta distribution fit.beta_pre_2017_NG <- fitdist(df_2017_NG$score_pre, "beta", method = "mme") fit.beta_post_2017_NG <- fitdist(df_2017_NG$score_post, "beta", method = "mme") plot(fit.beta_pre_2017_NG) plot(fit.beta_post_2017_NG) ############################################################################################ #Beta regression models #Function to transform y values to be used on a betareg distribution y.transf.betareg <- function(y){ n.obs <- sum(!is.na(y)) (y * (n.obs - 1) + 0.5) / n.obs } #Are the pre-recitation scores different across years (use of Genie or not)? effect_combined_years_pre <- betareg(y.transf.betareg(score_pre) ~ year, data=combined_years, link = "logit") summary(effect_combined_years_pre) #Are the post-recitation scores across years dependent of the pre-recitation scores and/or the year of instruction (use of Genie or not)? effect_combined_years <- betareg(y.transf.betareg(score_post) ~ score_pre + year, data=combined_years, link = "logit") summary(effect_combined_years) #Are the post-recitation scores on 2016 dependent of the pre-recitation scores and/or the class section? effect_test_2016 <- betareg(y.transf.betareg(score_post) ~ score_pre + class, data=df_2016, link = "logit") summary(effect_test_2016) #Are the post-recitation scores on 2017 dependent of the pre-recitation scores and/or the class section when using Genie? effect_test_2017_G <- betareg(y.transf.betareg(score_post) ~ score_pre + class, data=df_2017_G, link = "logit") summary(effect_test_2017_G) #Are the post-recitation scores on 2017 dependent of the pre-recitation scores and/or the class section when not using Genie? effect_test_2017_NG <- betareg(y.transf.betareg(score_post) ~ score_pre + class, data=df_2017_NG, link = "logit") summary(effect_test_2017_NG) #Writing output to text sink('T3_betareg_output.txt', append=TRUE) cat("Betareg on pre-recitation scores across years\n") summary(effect_combined_years_pre) cat("\n\n") cat("=============================\n") cat("Betareg on pre-recitation scores across years\n") summary(effect_combined_years) cat("\n\n") cat("=============================\n") cat("Betareg on postrecitation scores 2016\n") summary(effect_test_2016) cat("\n\n") cat("=============================\n") cat("Betareg on postrecitation scores 2017 Genie\n") summary(effect_test_2017_G) cat("\n\n") cat("=============================\n") cat("Betareg on postrecitation scores 2017 Non-genie\n") summary(effect_test_2017_NG) sink() ############################################################################################ #Size effect testing #Transforms the scores columns on each data frame into a numeric vector score_post_2016 <- as.numeric(as.character(df_2016$score_post)) score_pre_2016 <- as.numeric(as.character(df_2016$score_pre)) score_diff_2016 <- as.numeric(as.character(df_2016$diff)) score_post_2017_G <- as.numeric(as.character(df_2017_G$score_post)) score_pre_2017_G <- as.numeric(as.character(df_2017_G$score_pre)) score_diff_2017_G <- as.numeric(as.character(df_2017_G$diff)) score_post_2017_NG <- as.numeric(as.character(df_2017_NG$score_post)) score_pre_2017_NG <- as.numeric(as.character(df_2017_NG$score_pre)) score_diff_2017_NG <- as.numeric(as.character(df_2017_NG$diff)) #Are the scores different before and after instruction on 2016 (Genie)? cohens_prepost_2016 <- cohen.d(score_post_2016,score_pre_2016) #Are the scores different before and after instruction on 2017 (Genie)? cohens_prepost_2017G <- cohen.d(score_post_2017_G,score_pre_2017_G) #Are the scores different before and after instruction on 2017 (No Genie)? cohens_prepost_2017NG <- cohen.d(score_post_2017_NG,score_pre_2017_NG) #Are the pre-recitation scores different between 2017's Genie and Non Genie classes? cohens_pre_2017GNG <- cohen.d(score_pre_2017_G,score_pre_2017_NG) #Are the post-recitation scores different between 2017's Genie and Non Genie classes? cohens_post_2017GNG <- cohen.d(score_post_2017_G,score_post_2017_NG) #Are the difference in pre- and post-recitation scores different between 2017's Genie and Non Genie classes? cohens_diff_2017GNG <- cohen.d(score_diff_2017_G,score_diff_2017_NG) mean(score_pre_2016) mean(score_post_2016) mean(score_pre_2017_G) mean(score_post_2017_G) mean(score_pre_2017_NG) mean(score_post_2017_NG) #Writing output to text sink('T4_cohenD_output.txt', append=TRUE) cat("Cohen's d for pre- vs. post-recitation 2016\n") cohens_prepost_2016 cat("=============================\n") cat("Cohen's d for pre- vs. post-recitation 2017 Genie\n") cohens_prepost_2017G cat("=============================\n") cat("Cohen's d for pre- vs. post-recitation 2017 Non-genie\n") cohens_prepost_2017NG cat("=============================\n") cat("Cohen's d for pre-recitation 2017 Genie vs. Non-genie\n") cohens_pre_2017GNG cat("=============================\n") cat("Cohen's d for post-recitation 2017 Genie vs. Non-genie\n") cohens_post_2017GNG cat("=============================\n") cat("Cohen's d for diff between pre- and post-recitation 2017 Genie vs. Non-genie\n") cohens_diff_2017GNG sink() ############################################################################################ #Paired Student t-test of pre- vs post-recitation performance #2016 t.test(score_pre_2016, score_post_2016, paired = TRUE, p.adj = "bonf") #2017G t.test(score_pre_2017_G, score_post_2017_G, paired = TRUE, p.adj = "bonf") #2017NG t.test(score_pre_2017_NG, score_post_2017_NG, paired = TRUE, p.adj = "bonf") #Writing output to text sink('T5_paired_t_test.txt', append=TRUE) cat("Paired Student t-test of pre- vs post-recitation performance in 2016\n") t.test(score_pre_2016, score_post_2016, paired = TRUE, p.adj = "bonf") cat("=============================\n") cat("Paired Student t-test of pre- vs post-recitation performance in 2017 Genie\n") t.test(score_pre_2017_G, score_post_2017_G, paired = TRUE, p.adj = "bonf") cat("=============================\n") cat("Paired Student t-test of pre- vs post-recitation performance in 2017 Non-genie\n") t.test(score_pre_2017_NG, score_post_2017_NG, paired = TRUE, p.adj = "bonf") sink() ############################################################################################ #Visualization of performance on each year via violin plots VIOLINPLOT_2016_PRE_POST <- read.table("Sfile11_VIOLINPLOT_2016_PRE_POST.txt", header = TRUE, sep = "\t") head(VIOLINPLOT_2016_PRE_POST) VIOLINPLOT_2017_GENIE_PRE_POST <- read.table("Sfile12_VIOLINPLOT_2017_GENIE_PRE_POST.txt", header = TRUE, sep = "\t") head(VIOLINPLOT_2017_GENIE_PRE_POST) VIOLINPLOT_2017_NOGENIE_PRE_POST <- read.table("Sfile13_VIOLINPLOT_2017_NOGENIE_PRE_POST.txt", header = TRUE, sep = "\t") head(VIOLINPLOT_2017_NOGENIE_PRE_POST) require(gridExtra) VIOLINPLOT_2016 <- ggplot(VIOLINPLOT_2016_PRE_POST, aes(x = year, y = score, fill=year)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2016" = "Pre-recitation", "Post-recitation_score_2016" = "Post-recitation")) + xlab("") + labs(title="Genie 2016", y ="Scores")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + scale_fill_brewer(palette="Dark2") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") VIOLINPLOT_2017_GENIE <- ggplot(VIOLINPLOT_2017_GENIE_PRE_POST, aes(x = year, y = score, fill=year)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2017" = "Pre-recitation", "Post-recitation_score_2017" = "Post-recitation")) + xlab("") + labs(title="Genie 2017", y ="Scores")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + scale_fill_brewer(palette="Dark2") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") VIOLINPLOT_2017_NOGENIE <- ggplot(VIOLINPLOT_2017_NOGENIE_PRE_POST, aes(x = year, y = score, fill=year)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2017" = "Pre-recitation", "Post-recitation_score_2017" = "Post-recitation")) + xlab("") + labs(title="Non-Genie 2017", y ="Scores")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + scale_fill_brewer(palette="Dark2") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") #Paired student plot pd <- position_dodge(0.01) PAIRED_STUDENT_2016 <- ggplot(VIOLINPLOT_2016_PRE_POST, aes(x = year, y = score)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2016" = "Pre-recitation", "Post-recitation_score_2016" = "Post-recitation")) + xlab("") + labs(title="Genie 2016", y ="Scores")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + geom_line(aes(group = student, color=student), position = pd) + geom_point(position = pd) PAIRED_STUDENT_2017_GENIE <- ggplot(VIOLINPLOT_2017_GENIE_PRE_POST, aes(x = year, y = score)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2017" = "Pre-recitation", "Post-recitation_score_2017" = "Post-recitation")) + xlab("") + labs(title="Genie 2017", y ="")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + geom_line(aes(group = student, color=student), position = pd) + geom_point(position = pd) PAIRED_STUDENT_2017_NOGENIE <- ggplot(VIOLINPLOT_2017_NOGENIE_PRE_POST, aes(x = year, y = score)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2017" = "Pre-recitation", "Post-recitation_score_2017" = "Post-recitation")) + xlab("") + labs(title="Non-Genie 2017", y ="") + theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + geom_line(aes(group = student, color=student), position = pd) + geom_point(position = pd) pdf(file="Sfile14_violinplots.pdf") FS14 <- ggarrange(PAIRED_STUDENT_2016, PAIRED_STUDENT_2017_GENIE, PAIRED_STUDENT_2017_NOGENIE, ncol=3) print(FS14) dev.off() ############################################################################################ #Testing performance per question rawdat <- read_tsv("Sfile10_RESULTS_COMPILED_BY_QUESTION.txt") # Calculate McNemar's Test Statistic # https://en.wikipedia.org/wiki/McNemar%27s_test mcnemar_stat <- function(b,c) { (b-c)^2/(b+c) } # Reshape the input data so that each year-student-question has its own line. # Each observation will have a Pre and Post Stat dat <- rawdat %>% pivot_longer(!c(year,class,student)) %>% separate("name", c("quiz", "question"), sep="_") %>% mutate(question = as.integer(question)) %>% pivot_wider(names_from="quiz",values_from="value") # Count for each year-question the interaction of pre and post. # use these counts to measure McNemar's statistic and its pvalue. datn1 <- dat %>% mutate(result = glue("ans_{Pre}_{Post}")) %>% count(year, question, result) %>% pivot_wider(names_from="result",values_from="n", values_fill=list(n=0L)) %>% mutate(mcnemar_stat = mcnemar_stat(ans_1_0, ans_0_1), mcnemar_pvalue = pchisq(mcnemar_stat, 1L, lower.tail = FALSE)) # Save Results write_tsv(datn1,"T6_mcnemar.tsv") perQ_scores_2016_pval <- c(0.182, 2.88e-09, 0.086, 4.02e-05, 0.14, 2.79e-15, 0.075, 0.069, 9.37e-06, 6.15e-05, 0.004, 0.09, 0.317, 2.71e-05, 3.84e-08, 0.004, 0.113, 0.729, 1, 5.54e-05, 5.56e-04, 0.579) perQ_scores_G2017_pval <- c(0.317, 0.012, 0.072, 0.353, 0.002, 9.05e-06, 0.077, 0.04, 3.96e-05, 5.73e-09, 3.83e-05, 0.063, 0.004, 3.96e-05, 2.21e-07, 2.56e-04, 3.59e-04, 0.853, 0.013, 0.003, 0.018, 0.005) perQ_scores_NG2017_pval <- c(0.059, 0.102, 0.007, 0.549, 0.003, 2.78e-08, 0.691, 0.024, 9.43e-08, 8.98e-12, 5.97e-07, 1, 0.006, 6.07e-04, 1.86e-06, 0.012, 0.003, 0.014, 0.002, 9.05e-06, 0.033, 0.273) p.adjust(perQ_scores_2016_pval, "bonferroni", n = length(perQ_scores_2016_pval)) p.adjust(perQ_scores_G2017_pval, "bonferroni", n = length(perQ_scores_G2017_pval)) p.adjust(perQ_scores_NG2017_pval, "bonferroni", n = length(perQ_scores_NG2017_pval)) ############################################################################################ #Comparing 2017 genie vs non-genie classes performance per question rawdat <- read_tsv("Sfile10_RESULTS_COMPILED_BY_QUESTION.txt") dat0 <- rawdat %>% pivot_longer(!c(year,class,student)) %>% separate("name", c("quiz", "question"), sep="_") %>% mutate(question = as.integer(question)) %>% pivot_wider(names_from="quiz",values_from="value") write_tsv(dat0,"pre_and_post.tsv") dat <- read_tsv("pre_and_post.tsv") # Compare 2017 Genie and 2017 Not Genie on pre and post # Use Fisher's exact tests to test the association between question switching and # type of instruction. dat1 <- dat %>% mutate(result = str_glue("ans_{Pre}_{Post}")) %>% count(year, question, result) %>% pivot_wider(names_from="result",values_from="n", values_fill=list(n=0L)) %>% filter( year %in% c("2017_Genie", "2017_No_Genie")) dat2 <- dat1 %>% select(year,question,ans_0_1,ans_1_0) %>% pivot_wider(names_from="year", values_from=c(ans_0_1,ans_1_0)) f <- function(a,b,c,d) { m <- matrix(c(a,b,c,d),nrow=2) v <- fisher.test(m) data.frame(OR.est=v$estimate, p.value=v$p.value) } dat3 <- dat2 %>% rowwise() %>% summarize(f(ans_0_1_2017_Genie, ans_0_1_2017_No_Genie, ans_1_0_2017_Genie, ans_1_0_2017_No_Genie)) dat4 <- bind_cols(dat2,dat3) write_tsv(dat4,"T7_genie_vs_non.tsv") pal_okabe_ito <- c("#E69F00", "#56B4E9", "#009E73", "#F0E442", "#0072B2", "#D55E00", "#CC79A7", "#000000") dat <- read_tsv("pre_and_post.tsv") dat.A <- dat %>% mutate(result = str_glue("ans_{Pre}_{Post}")) %>% count(year, student, result) %>% pivot_wider(names_from="result",values_from="n", values_fill=list(n=0L)) %>% mutate(pre_acc = (ans_1_1+ans_1_0), post_acc = (ans_1_1+ans_0_1)) dat.B <- dat.A %>% mutate(diff = post_acc - pre_acc, diffs = sign(diff), excess = "no") %>% arrange(diff) dat4 <- dat.B %>% group_by(year) %>% group_modify(function(tbl,y){ j <- tail(1:nrow(tbl),sum(tbl$diffs)) tbl$excess[j] <- "yes" tbl }) pretty_year <- c( `2016_Genie` = "Genie 2016", `2017_Genie` = "Genie 2017", `2017_No_Genie` = "Non-Genie 2017" ) ann_tbl <- tibble(pre_acc=10,post_acc=2.5,year="2017_No_Genie",excess="yes") pdf(file="F1.pdf") set.seed(0x90210) gg <- ggplot(dat4, aes(x=pre_acc,y=post_acc,color=excess)) + geom_jitter(width=0.3,height=0.3) gg <- gg + geom_abline(slope=1,intercept=0) gg <- gg + labs(x = "Pre-Score", y = "Post-Score") gg <- gg + theme_minimal() + coord_fixed(xlim=c(0,22),ylim=c(0,22)) gg <- gg + facet_wrap(~year, labeller = labeller(year = pretty_year)) gg <- gg + scale_color_manual(values=pal_okabe_ito[c(8,2)]) gg <- gg + theme(legend.position = "none") gg <- gg + geom_label(data=ann_tbl,label=" Excess\n Improvement",hjust="left", nudge_x=-2,color="black",size=3) + geom_point(data=ann_tbl,size=5) print(gg) dev.off() ############################################################################################ compiled_years_by_question <- read.table("Sfile10_RESULTS_COMPILED_BY_QUESTION.txt", header = TRUE, sep = "\t") split_years_by_year <- split(compiled_years_by_question, compiled_years_by_question$year) df_2016_by_year <- split_years_by_year[[1]] df_2017_G_by_year <- split_years_by_year[[2]] df_2017_NG_by_year <- split_years_by_year[[3]] #Transforms the pre-recitation scores columns on each data frame into a numeric vector score_pre1_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_1)) score_pre2_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_2)) score_pre3_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_3)) score_pre4_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_4)) score_pre5_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_5)) score_pre6_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_6)) score_pre7_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_7)) score_pre8_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_8)) score_pre9_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_9)) score_pre10_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_10)) score_pre11_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_11)) score_pre12_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_12)) score_pre13_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_13)) score_pre14_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_14)) score_pre15_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_15)) score_pre16_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_16)) score_pre17_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_17)) score_pre18_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_18)) score_pre19_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_19)) score_pre20_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_20)) score_pre21_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_21)) score_pre22_2017_G <- as.numeric(as.character(df_2017_G_by_year$Pre_22)) score_pre1_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_1)) score_pre2_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_2)) score_pre3_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_3)) score_pre4_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_4)) score_pre5_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_5)) score_pre6_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_6)) score_pre7_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_7)) score_pre8_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_8)) score_pre9_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_9)) score_pre10_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_10)) score_pre11_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_11)) score_pre12_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_12)) score_pre13_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_13)) score_pre14_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_14)) score_pre15_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_15)) score_pre16_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_16)) score_pre17_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_17)) score_pre18_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_18)) score_pre19_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_19)) score_pre20_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_20)) score_pre21_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_21)) score_pre22_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Pre_22)) #Gets score for each question score_pre1_2017_G_val <- sum(score_pre1_2017_G)/length(score_pre1_2017_G) score_pre2_2017_G_val <- sum(score_pre2_2017_G)/length(score_pre2_2017_G) score_pre3_2017_G_val <- sum(score_pre3_2017_G)/length(score_pre3_2017_G) score_pre4_2017_G_val <- sum(score_pre4_2017_G)/length(score_pre4_2017_G) score_pre5_2017_G_val <- sum(score_pre5_2017_G)/length(score_pre5_2017_G) score_pre6_2017_G_val <- sum(score_pre6_2017_G)/length(score_pre6_2017_G) score_pre7_2017_G_val <- sum(score_pre7_2017_G)/length(score_pre7_2017_G) score_pre8_2017_G_val <- sum(score_pre8_2017_G)/length(score_pre8_2017_G) score_pre9_2017_G_val <- sum(score_pre9_2017_G)/length(score_pre9_2017_G) score_pre10_2017_G_val <- sum(score_pre10_2017_G)/length(score_pre10_2017_G) score_pre11_2017_G_val <- sum(score_pre11_2017_G)/length(score_pre11_2017_G) score_pre12_2017_G_val <- sum(score_pre12_2017_G)/length(score_pre12_2017_G) score_pre13_2017_G_val <- sum(score_pre13_2017_G)/length(score_pre13_2017_G) score_pre14_2017_G_val <- sum(score_pre14_2017_G)/length(score_pre14_2017_G) score_pre15_2017_G_val <- sum(score_pre15_2017_G)/length(score_pre15_2017_G) score_pre16_2017_G_val <- sum(score_pre16_2017_G)/length(score_pre16_2017_G) score_pre17_2017_G_val <- sum(score_pre17_2017_G)/length(score_pre17_2017_G) score_pre18_2017_G_val <- sum(score_pre18_2017_G)/length(score_pre18_2017_G) score_pre19_2017_G_val <- sum(score_pre19_2017_G)/length(score_pre19_2017_G) score_pre20_2017_G_val <- sum(score_pre20_2017_G)/length(score_pre20_2017_G) score_pre21_2017_G_val <- sum(score_pre21_2017_G)/length(score_pre21_2017_G) score_pre22_2017_G_val <- sum(score_pre22_2017_G)/length(score_pre22_2017_G) score_pre1_2017_NG_val <- sum(score_pre1_2017_NG)/length(score_pre1_2017_NG) score_pre2_2017_NG_val <- sum(score_pre2_2017_NG)/length(score_pre2_2017_NG) score_pre3_2017_NG_val <- sum(score_pre3_2017_NG)/length(score_pre3_2017_NG) score_pre4_2017_NG_val <- sum(score_pre4_2017_NG)/length(score_pre4_2017_NG) score_pre5_2017_NG_val <- sum(score_pre5_2017_NG)/length(score_pre5_2017_NG) score_pre6_2017_NG_val <- sum(score_pre6_2017_NG)/length(score_pre6_2017_NG) score_pre7_2017_NG_val <- sum(score_pre7_2017_NG)/length(score_pre7_2017_NG) score_pre8_2017_NG_val <- sum(score_pre8_2017_NG)/length(score_pre8_2017_NG) score_pre9_2017_NG_val <- sum(score_pre9_2017_NG)/length(score_pre9_2017_NG) score_pre10_2017_NG_val <- sum(score_pre10_2017_NG)/length(score_pre10_2017_NG) score_pre11_2017_NG_val <- sum(score_pre11_2017_NG)/length(score_pre11_2017_NG) score_pre12_2017_NG_val <- sum(score_pre12_2017_NG)/length(score_pre12_2017_NG) score_pre13_2017_NG_val <- sum(score_pre13_2017_NG)/length(score_pre13_2017_NG) score_pre14_2017_NG_val <- sum(score_pre14_2017_NG)/length(score_pre14_2017_NG) score_pre15_2017_NG_val <- sum(score_pre15_2017_NG)/length(score_pre15_2017_NG) score_pre16_2017_NG_val <- sum(score_pre16_2017_NG)/length(score_pre16_2017_NG) score_pre17_2017_NG_val <- sum(score_pre17_2017_NG)/length(score_pre17_2017_NG) score_pre18_2017_NG_val <- sum(score_pre18_2017_NG)/length(score_pre18_2017_NG) score_pre19_2017_NG_val <- sum(score_pre19_2017_NG)/length(score_pre19_2017_NG) score_pre20_2017_NG_val <- sum(score_pre20_2017_NG)/length(score_pre20_2017_NG) score_pre21_2017_NG_val <- sum(score_pre21_2017_NG)/length(score_pre21_2017_NG) score_pre22_2017_NG_val <- sum(score_pre22_2017_NG)/length(score_pre22_2017_NG) #Transforms the post-recitation scores columns on each data frame into a numeric vector score_post1_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_1)) score_post2_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_2)) score_post3_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_3)) score_post4_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_4)) score_post5_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_5)) score_post6_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_6)) score_post7_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_7)) score_post8_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_8)) score_post9_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_9)) score_post10_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_10)) score_post11_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_11)) score_post12_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_12)) score_post13_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_13)) score_post14_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_14)) score_post15_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_15)) score_post16_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_16)) score_post17_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_17)) score_post18_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_18)) score_post19_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_19)) score_post20_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_20)) score_post21_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_21)) score_post22_2017_G <- as.numeric(as.character(df_2017_G_by_year$Post_22)) score_post1_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_1)) score_post2_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_2)) score_post3_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_3)) score_post4_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_4)) score_post5_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_5)) score_post6_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_6)) score_post7_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_7)) score_post8_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_8)) score_post9_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_9)) score_post10_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_10)) score_post11_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_11)) score_post12_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_12)) score_post13_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_13)) score_post14_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_14)) score_post15_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_15)) score_post16_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_16)) score_post17_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_17)) score_post18_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_18)) score_post19_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_19)) score_post20_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_20)) score_post21_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_21)) score_post22_2017_NG <- as.numeric(as.character(df_2017_NG_by_year$Post_22)) #Gets score for each question score_post1_2017_G_val <- sum(score_post1_2017_G)/length(score_post1_2017_G) score_post2_2017_G_val <- sum(score_post2_2017_G)/length(score_post2_2017_G) score_post3_2017_G_val <- sum(score_post3_2017_G)/length(score_post3_2017_G) score_post4_2017_G_val <- sum(score_post4_2017_G)/length(score_post4_2017_G) score_post5_2017_G_val <- sum(score_post5_2017_G)/length(score_post5_2017_G) score_post6_2017_G_val <- sum(score_post6_2017_G)/length(score_post6_2017_G) score_post7_2017_G_val <- sum(score_post7_2017_G)/length(score_post7_2017_G) score_post8_2017_G_val <- sum(score_post8_2017_G)/length(score_post8_2017_G) score_post9_2017_G_val <- sum(score_post9_2017_G)/length(score_post9_2017_G) score_post10_2017_G_val <- sum(score_post10_2017_G)/length(score_post10_2017_G) score_post11_2017_G_val <- sum(score_post11_2017_G)/length(score_post11_2017_G) score_post12_2017_G_val <- sum(score_post12_2017_G)/length(score_post12_2017_G) score_post13_2017_G_val <- sum(score_post13_2017_G)/length(score_post13_2017_G) score_post14_2017_G_val <- sum(score_post14_2017_G)/length(score_post14_2017_G) score_post15_2017_G_val <- sum(score_post15_2017_G)/length(score_post15_2017_G) score_post16_2017_G_val <- sum(score_post16_2017_G)/length(score_post16_2017_G) score_post17_2017_G_val <- sum(score_post17_2017_G)/length(score_post17_2017_G) score_post18_2017_G_val <- sum(score_post18_2017_G)/length(score_post18_2017_G) score_post19_2017_G_val <- sum(score_post19_2017_G)/length(score_post19_2017_G) score_post20_2017_G_val <- sum(score_post20_2017_G)/length(score_post20_2017_G) score_post21_2017_G_val <- sum(score_post21_2017_G)/length(score_post21_2017_G) score_post22_2017_G_val <- sum(score_post22_2017_G)/length(score_post22_2017_G) score_post1_2017_NG_val <- sum(score_post1_2017_NG)/length(score_post1_2017_NG) score_post2_2017_NG_val <- sum(score_post2_2017_NG)/length(score_post2_2017_NG) score_post3_2017_NG_val <- sum(score_post3_2017_NG)/length(score_post3_2017_NG) score_post4_2017_NG_val <- sum(score_post4_2017_NG)/length(score_post4_2017_NG) score_post5_2017_NG_val <- sum(score_post5_2017_NG)/length(score_post5_2017_NG) score_post6_2017_NG_val <- sum(score_post6_2017_NG)/length(score_post6_2017_NG) score_post7_2017_NG_val <- sum(score_post7_2017_NG)/length(score_post7_2017_NG) score_post8_2017_NG_val <- sum(score_post8_2017_NG)/length(score_post8_2017_NG) score_post9_2017_NG_val <- sum(score_post9_2017_NG)/length(score_post9_2017_NG) score_post10_2017_NG_val <- sum(score_post10_2017_NG)/length(score_post10_2017_NG) score_post11_2017_NG_val <- sum(score_post11_2017_NG)/length(score_post11_2017_NG) score_post12_2017_NG_val <- sum(score_post12_2017_NG)/length(score_post12_2017_NG) score_post13_2017_NG_val <- sum(score_post13_2017_NG)/length(score_post13_2017_NG) score_post14_2017_NG_val <- sum(score_post14_2017_NG)/length(score_post14_2017_NG) score_post15_2017_NG_val <- sum(score_post15_2017_NG)/length(score_post15_2017_NG) score_post16_2017_NG_val <- sum(score_post16_2017_NG)/length(score_post16_2017_NG) score_post17_2017_NG_val <- sum(score_post17_2017_NG)/length(score_post17_2017_NG) score_post18_2017_NG_val <- sum(score_post18_2017_NG)/length(score_post18_2017_NG) score_post19_2017_NG_val <- sum(score_post19_2017_NG)/length(score_post19_2017_NG) score_post20_2017_NG_val <- sum(score_post20_2017_NG)/length(score_post20_2017_NG) score_post21_2017_NG_val <- sum(score_post21_2017_NG)/length(score_post21_2017_NG) score_post22_2017_NG_val <- sum(score_post22_2017_NG)/length(score_post22_2017_NG) #Vector of pre-recitation and post-recitation scores by question on 2017 genie classes pre_recitation_scores_2017_G2AT3 <- c(score_pre1_2017_G_val, score_pre3_2017_G_val, score_pre15_2017_G_val, score_pre4_2017_G_val, score_pre10_2017_G_val, score_pre13_2017_G_val, score_pre16_2017_G_val, score_pre7_2017_G_val, score_pre5_2017_G_val, score_pre6_2017_G_val, score_pre8_2017_G_val, score_pre2_2017_G_val, score_pre9_2017_G_val, score_pre12_2017_G_val, score_pre17_2017_G_val, score_pre20_2017_G_val, score_pre14_2017_G_val, score_pre19_2017_G_val, score_pre22_2017_G_val, score_pre11_2017_G_val, score_pre18_2017_G_val, score_pre21_2017_G_val) post_recitation_scores_2017_G2AT3 <- c(score_post1_2017_G_val, score_post3_2017_G_val, score_post15_2017_G_val, score_post4_2017_G_val, score_post10_2017_G_val, score_post13_2017_G_val, score_post16_2017_G_val, score_post7_2017_G_val, score_post5_2017_G_val, score_post6_2017_G_val, score_post8_2017_G_val, score_post2_2017_G_val, score_post9_2017_G_val, score_post12_2017_G_val, score_post17_2017_G_val, score_post20_2017_G_val, score_post14_2017_G_val, score_post19_2017_G_val, score_post22_2017_G_val, score_post11_2017_G_val, score_post18_2017_G_val, score_post21_2017_G_val) pre_recitation_scores_2017_NG2AT3 <- c(score_pre1_2017_NG_val, score_pre3_2017_NG_val, score_pre15_2017_NG_val, score_pre4_2017_NG_val, score_pre10_2017_NG_val, score_pre13_2017_NG_val, score_pre16_2017_NG_val, score_pre7_2017_NG_val, score_pre5_2017_NG_val, score_pre6_2017_NG_val, score_pre8_2017_NG_val, score_pre2_2017_NG_val, score_pre9_2017_NG_val, score_pre12_2017_NG_val, score_pre17_2017_NG_val, score_pre20_2017_NG_val, score_pre14_2017_NG_val, score_pre19_2017_NG_val, score_pre22_2017_NG_val, score_pre11_2017_NG_val, score_pre18_2017_NG_val, score_pre21_2017_NG_val) post_recitation_scores_2017_NG2AT3 <- c(score_post1_2017_NG_val, score_post3_2017_NG_val, score_post15_2017_NG_val, score_post4_2017_NG_val, score_post10_2017_NG_val, score_post13_2017_NG_val, score_post16_2017_NG_val, score_post7_2017_NG_val, score_post5_2017_NG_val, score_post6_2017_NG_val, score_post8_2017_NG_val, score_post2_2017_NG_val, score_post9_2017_NG_val, score_post12_2017_NG_val, score_post17_2017_NG_val, score_post20_2017_NG_val, score_post14_2017_NG_val, score_post19_2017_NG_val, score_post22_2017_NG_val, score_post11_2017_NG_val, score_post18_2017_NG_val, score_post21_2017_NG_val) #get diifferece between pre- and post-recitation scores score_diff1_2017_G <- score_post1_2017_G_val-score_pre1_2017_G_val score_diff2_2017_G <- score_post2_2017_G_val-score_pre2_2017_G_val score_diff3_2017_G <- score_post3_2017_G_val-score_pre3_2017_G_val score_diff4_2017_G <- score_post4_2017_G_val-score_pre4_2017_G_val score_diff5_2017_G <- score_post5_2017_G_val-score_pre5_2017_G_val score_diff6_2017_G <- score_post6_2017_G_val-score_pre6_2017_G_val score_diff7_2017_G <- score_post7_2017_G_val-score_pre7_2017_G_val score_diff8_2017_G <- score_post8_2017_G_val-score_pre8_2017_G_val score_diff9_2017_G <- score_post9_2017_G_val-score_pre9_2017_G_val score_diff10_2017_G <- score_post10_2017_G_val-score_pre10_2017_G_val score_diff11_2017_G <- score_post11_2017_G_val-score_pre11_2017_G_val score_diff12_2017_G <- score_post12_2017_G_val-score_pre12_2017_G_val score_diff13_2017_G <- score_post13_2017_G_val-score_pre13_2017_G_val score_diff14_2017_G <- score_post14_2017_G_val-score_pre14_2017_G_val score_diff15_2017_G <- score_post15_2017_G_val-score_pre15_2017_G_val score_diff16_2017_G <- score_post16_2017_G_val-score_pre16_2017_G_val score_diff17_2017_G <- score_post17_2017_G_val-score_pre17_2017_G_val score_diff18_2017_G <- score_post18_2017_G_val-score_pre18_2017_G_val score_diff19_2017_G <- score_post19_2017_G_val-score_pre19_2017_G_val score_diff20_2017_G <- score_post20_2017_G_val-score_pre20_2017_G_val score_diff21_2017_G <- score_post21_2017_G_val-score_pre21_2017_G_val score_diff22_2017_G <- score_post22_2017_G_val-score_pre22_2017_G_val diff_recitation_scores_2017_G2AT3 <- c(score_diff1_2017_G, score_diff3_2017_G, score_diff4_2017_G, score_diff10_2017_G, score_diff13_2017_G, score_diff15_2017_G, score_diff16_2017_G, score_diff7_2017_G, score_diff5_2017_G, score_diff6_2017_G, score_diff8_2017_G, score_diff2_2017_G, score_diff9_2017_G, score_diff12_2017_G, score_diff17_2017_G, score_diff20_2017_G, score_diff14_2017_G, score_diff19_2017_G, score_diff22_2017_G, score_diff11_2017_G, score_diff18_2017_G, score_diff21_2017_G) score_diff1_2017_NG <- score_post1_2017_NG_val-score_pre1_2017_NG_val score_diff2_2017_NG <- score_post2_2017_NG_val-score_pre2_2017_NG_val score_diff3_2017_NG <- score_post3_2017_NG_val-score_pre3_2017_NG_val score_diff4_2017_NG <- score_post4_2017_NG_val-score_pre4_2017_NG_val score_diff5_2017_NG <- score_post5_2017_NG_val-score_pre5_2017_NG_val score_diff6_2017_NG <- score_post6_2017_NG_val-score_pre6_2017_NG_val score_diff7_2017_NG <- score_post7_2017_NG_val-score_pre7_2017_NG_val score_diff8_2017_NG <- score_post8_2017_NG_val-score_pre8_2017_NG_val score_diff9_2017_NG <- score_post9_2017_NG_val-score_pre9_2017_NG_val score_diff10_2017_NG <- score_post10_2017_NG_val-score_pre10_2017_NG_val score_diff11_2017_NG <- score_post11_2017_NG_val-score_pre11_2017_NG_val score_diff12_2017_NG <- score_post12_2017_NG_val-score_pre12_2017_NG_val score_diff13_2017_NG <- score_post13_2017_NG_val-score_pre13_2017_NG_val score_diff14_2017_NG <- score_post14_2017_NG_val-score_pre14_2017_NG_val score_diff15_2017_NG <- score_post15_2017_NG_val-score_pre15_2017_NG_val score_diff16_2017_NG <- score_post16_2017_NG_val-score_pre16_2017_NG_val score_diff17_2017_NG <- score_post17_2017_NG_val-score_pre17_2017_NG_val score_diff18_2017_NG <- score_post18_2017_NG_val-score_pre18_2017_NG_val score_diff19_2017_NG <- score_post19_2017_NG_val-score_pre19_2017_NG_val score_diff20_2017_NG <- score_post20_2017_NG_val-score_pre20_2017_NG_val score_diff21_2017_NG <- score_post21_2017_NG_val-score_pre21_2017_NG_val score_diff22_2017_NG <- score_post22_2017_NG_val-score_pre22_2017_NG_val diff_recitation_scores_2017_NG2AT3 <- c(score_diff1_2017_NG, score_diff3_2017_NG, score_diff4_2017_NG, score_diff10_2017_NG, score_diff13_2017_NG, score_diff15_2017_NG, score_diff16_2017_NG, score_diff7_2017_NG, score_diff5_2017_NG, score_diff6_2017_NG, score_diff8_2017_NG, score_diff2_2017_NG, score_diff9_2017_NG, score_diff12_2017_NG, score_diff17_2017_NG, score_diff20_2017_NG, score_diff14_2017_NG, score_diff19_2017_NG, score_diff22_2017_NG, score_diff11_2017_NG, score_diff18_2017_NG, score_diff21_2017_NG) #Questions on 2AT3 order in Price et al 2014 question2AT3 <- c("Q1","Q3","Q15","Q4","Q10","Q13","Q16","Q7","Q5","Q6","Q8","Q2","Q9","Q12","Q17","Q20","Q14","Q19","Q22","Q11","Q18","Q21") time_diffG <- rep("Diff-score Genie", 22) time_diffNG <- rep("Diff-score no Genie", 22) df_diff_scores_by_question_GA2T3 <- data.frame(diff_recitation_scores = diff_recitation_scores_2017_G2AT3, question=question2AT3, time=time_diffG) df_diff_scores_by_question_NGA2T3 <- data.frame(diff_recitation_scores = diff_recitation_scores_2017_NG2AT3, question=question2AT3, time=time_diffNG) df_diff_scores_by_question_GA2T3$question <- factor(df_diff_scores_by_question_GA2T3$question, levels = df_diff_scores_by_question_GA2T3$question) df_diff_scores_by_question_NGA2T3$question <- factor(df_diff_scores_by_question_NGA2T3$question, levels = df_diff_scores_by_question_NGA2T3$question) df_diff_scores_by_question_combinedA2T3 <- rbind(df_diff_scores_by_question_GA2T3, df_diff_scores_by_question_NGA2T3) pdf(file="F3.pdf") my_purple_colors <- c("orchid4","plum") diff_per_question_plotA2T3 <- ggplot(data=df_diff_scores_by_question_combinedA2T3, aes(x=question, y=diff_recitation_scores, fill=time)) + geom_bar(width=0.4, stat="identity", position=position_dodge(0.5)) + scale_fill_manual(values = my_purple_colors) + xlab("Question") + ylab("Scores") + labs(fill = "Recitation") + scale_x_discrete(limits= df_diff_scores_by_question_combinedA2T3$p.question) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) print(diff_per_question_plotA2T3) dev.off() question2AT3 <- c("Q1","Q3","Q15","Q4","Q10","Q13","Q16","Q7","Q5","Q6","Q8","Q2","Q9","Q12","Q17","Q20","Q14","Q19","Q22","Q11","Q18","Q21") time_postG <- rep("Post-score Genie", 22) time_postNG <- rep("Post-score no Genie", 22) df_post_scores_by_question_GA2T3 <- data.frame(post_recitation_scores = post_recitation_scores_2017_G2AT3, question=question2AT3, time=time_postG) df_post_scores_by_question_NGA2T3 <- data.frame(post_recitation_scores = post_recitation_scores_2017_NG2AT3, question=question2AT3, time=time_postNG) df_post_scores_by_question_GA2T3$question <- factor(df_post_scores_by_question_GA2T3$question, levels = df_post_scores_by_question_GA2T3$question) df_post_scores_by_question_NGA2T3$question <- factor(df_post_scores_by_question_NGA2T3$question, levels = df_post_scores_by_question_NGA2T3$question) df_post_scores_by_question_combinedA2T3 <- rbind(df_post_scores_by_question_GA2T3, df_post_scores_by_question_NGA2T3) pdf(file="F2.pdf") my_green_colors <- c("darkgreen","darkseagreen3") post_per_question_plotA2T3 <- ggplot(data=df_post_scores_by_question_combinedA2T3, aes(x=question, y=post_recitation_scores, fill=time)) + geom_bar(width=0.4, stat="identity", position=position_dodge(0.5)) + #geom_text(aes(label=combined_diff_2017), vjust=1.6, color="red", # position = position_dodge(0.9), size=3.5)+ scale_fill_manual(values = my_green_colors) + xlab("Question") + ylab("Scores") + labs(fill = "Recitation") + scale_x_discrete(limits= df_post_scores_by_question_combinedA2T3$p.question)+ theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) print(post_per_question_plotA2T3) dev.off() ######################################################################################### ######################################################################################### ######################################################################################### ######################################################################################### ######################################################################################### #ANALYSIS REMOVING HONOR SECTIONS #The results for both years and treatments removing the honors recitation section on each year combined_years_without_honors <- combined_years[!(combined_years$class=="Pair1_730" & combined_years$year=="2016_Genie"),] combined_years_without_honors <- combined_years_without_honors[!(combined_years_without_honors$class=="Pair2_130" & combined_years_without_honors$year=="2017_Genie"),] head(combined_years_without_honors) #Creates a violin plot of the difference in scores for each year and treatment plotdiff <- ggplot(combined_years_without_honors, aes(x = year, y = diff, fill=year)) + scale_y_continuous(limits = c(0, 1), breaks = seq(0, 1, by = 0.1)) + scale_x_discrete() + xlab("Treatment Group") + ylab("Difference in recitation scores") + theme(legend.position = "none") + scale_fill_brewer(palette="Paired") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") plotdiff <- plotdiff + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) plotdiff ############################################################################################ #Splits the data frame into years and treatments split_years <- split(combined_years_without_honors, combined_years_without_honors$year) df_2016_without_honors <- split_years[[1]] df_2017_G_without_honors <- split_years[[2]] df_2017_NG_without_honors <- split_years[[3]] #Fitting the data to a distribution descdist(df_2016$score_pre, discrete = F) #beta distributionn descdist(df_2016$score_post, discrete = F) #beta distribution fit.beta_pre_2016 <- fitdist(df_2016_without_honors$score_pre, "beta", method = "mme") fit.beta_post_2016 <- fitdist(df_2016_without_honors$score_post, "beta", method = "mme") plot(fit.beta_pre_2016) plot(fit.beta_post_2016) descdist(df_2017_G$score_pre, discrete = F) #beta distributionn descdist(df_2017_G$score_post, discrete = F) #beta distribution fit.beta_pre_2017_G <- fitdist(df_2017_G_without_honors$score_pre, "beta", method = "mme") fit.beta_post_2017_G <- fitdist(df_2017_G_without_honors$score_post, "beta", method = "mme") plot(fit.beta_pre_2017_G) plot(fit.beta_post_2017_G) descdist(df_2017_NG$score_pre, discrete = F) #beta distributionn descdist(df_2017_NG$score_post, discrete = F) #beta distribution fit.beta_pre_2017_NG <- fitdist(df_2017_NG_without_honors$score_pre, "beta", method = "mme") fit.beta_post_2017_NG <- fitdist(df_2017_NG_without_honors$score_post, "beta", method = "mme") plot(fit.beta_pre_2017_NG) plot(fit.beta_post_2017_NG) ######Beta regression models####### #Function to transform y values to be used on a betareg distribution y.transf.betareg <- function(y){ n.obs <- sum(!is.na(y)) (y * (n.obs - 1) + 0.5) / n.obs } #Are the pre-recitation scores different across years (use of Genie or not)? effect_combined_years_without_honors_pre <- betareg(y.transf.betareg(score_pre) ~ year, data=combined_years_without_honors, link = "logit") #Are the post-recitation scores across years dependent of the pre-recitation scores and/or the year of instruction (use of Genie or not)? effect_combined_years_without_honors <- betareg(y.transf.betareg(score_post) ~ score_pre + year, data=combined_years_without_honors, link = "logit") #Are the post-recitation scores on 2016 dependent of the pre-recitation scores and/or the class section? effect_test_2016_without_honors <- betareg(y.transf.betareg(score_post) ~ score_pre + class, data=df_2016_without_honors, link = "logit") #Are the post-recitation scores on 2017 dependent of the pre-recitation scores and/or the class section when using Genie? effect_test_2017_G_without_honors <- betareg(y.transf.betareg(score_post) ~ score_pre + class, data=df_2017_G_without_honors, link = "logit") #Are the post-recitation scores on 2017 dependent of the pre-recitation scores and/or the class section when not using Genie? effect_test_2017_NG_without_honors <- betareg(y.transf.betareg(score_post) ~ score_pre + class, data=df_2017_NG_without_honors, link = "logit") #Writing output to text sink('nohonors_betareg_output.txt', append=TRUE) cat("Betareg on pre-recitation scores across years nohonors\n") summary(effect_combined_years_without_honors_pre) cat("\n\n") cat("=============================\n") cat("Betareg on pre-recitation scores across years nohonors\n") summary(effect_combined_years_without_honors) cat("\n\n") cat("=============================\n") cat("Betareg on postrecitation scores 2016 nohonors\n") summary(effect_test_2016_without_honors) cat("\n\n") cat("=============================\n") cat("Betareg on postrecitation scores 2017 Genie nohonors\n") summary(effect_test_2017_G_without_honors) cat("\n\n") cat("=============================\n") cat("Betareg on postrecitation scores 2017 Non-genie nohonors\n") summary(effect_test_2017_NG_without_honors) sink() ############################################################################################ #Size effect testing #Transforms the scores columns on each data frame into a numeric vector score_post_2016_without_honors <- as.numeric(as.character(df_2016_without_honors$score_post)) score_pre_2016_without_honors <- as.numeric(as.character(df_2016_without_honors$score_pre)) score_diff_2016_without_honors <- as.numeric(as.character(df_2016_without_honors$diff)) score_post_2017_G_without_honors <- as.numeric(as.character(df_2017_G_without_honors$score_post)) score_pre_2017_G_without_honors <- as.numeric(as.character(df_2017_G_without_honors$score_pre)) score_diff_2017_G_without_honors <- as.numeric(as.character(df_2017_G_without_honors$diff)) score_post_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_without_honors$score_post)) score_pre_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_without_honors$score_pre)) score_diff_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_without_honors$diff)) #Are the scores different before and after instruction on 2016 (Genie)? cohen.d(score_post_2016_without_honors,score_pre_2016_without_honors) #Are the scores different before and after instruction on 2017 (Genie)? cohen.d(score_post_2017_G_without_honors,score_pre_2017_G_without_honors) #Are the scores different before and after instruction on 2017 (No Genie)? cohen.d(score_post_2017_NG_without_honors,score_pre_2017_NG_without_honors) #Are the pre-recitation scores different between 2017's Genie and Non Genie classes? cohen.d(score_pre_2017_G_without_honors,score_pre_2017_NG_without_honors) #Are the post-recitation scores different between 2017's Genie and Non Genie classes? cohen.d(score_post_2017_G_without_honors,score_post_2017_NG_without_honors) #Are the difference in pre- and post-recitation scores different between 2017's Genie and Non Genie classes? cohen.d(score_diff_2017_G_without_honors,score_diff_2017_NG_without_honors) #Writing output to text sink('nohonors_cohenD_output.txt', append=TRUE) cat("Cohen's d for pre- vs. post-recitation 2016 nohonors\n") cohen.d(score_post_2016_without_honors,score_pre_2016_without_honors) cat("=============================\n") cat("Cohen's d for pre- vs. post-recitation 2017 Genie nohonors\n") cohen.d(score_post_2017_G_without_honors,score_pre_2017_G_without_honors) cat("=============================\n") cat("Cohen's d for pre- vs. post-recitation 2017 Non-genie nohonors\n") cohen.d(score_post_2017_NG_without_honors,score_pre_2017_NG_without_honors) cat("=============================\n") cat("Cohen's d for pre-recitation 2017 Genie vs. Non-genie nohonors\n") cohen.d(score_pre_2017_G_without_honors,score_pre_2017_NG_without_honors) cat("=============================\n") cat("Cohen's d for post-recitation 2017 Genie vs. Non-genie nohonors\n") cohen.d(score_post_2017_G_without_honors,score_post_2017_NG_without_honors) cat("=============================\n") cat("Cohen's d for diff between pre- and post-recitation 2017 Genie vs. Non-genie nohonors\n") cohen.d(score_diff_2017_G_without_honors,score_diff_2017_NG_without_honors) sink() ############################################################################################ #Paired Student t-test of pre- vs post-recitation performance #2016 t.test(score_pre_2016_without_honors, score_post_2016_without_honors, paired = TRUE, p.adj = "bonf") #2017G t.test(score_pre_2017_G_without_honors, score_post_2017_G_without_honors, paired = TRUE, p.adj = "bonf") #2017NG t.test(score_pre_2017_NG_without_honors, score_post_2017_NG_without_honors, paired = TRUE, p.adj = "bonf") #Writing output to text sink('nohonors_paired_t_test.txt', append=TRUE) cat("Paired Student t-test of pre- vs post-recitation performance in 2016 nohonors\n") t.test(score_pre_2016_without_honors, score_post_2016_without_honors, paired = TRUE, p.adj = "bonf") cat("=============================\n") cat("Paired Student t-test of pre- vs post-recitation performance in 2017 Genie nohonors\n") t.test(score_pre_2017_G_without_honors, score_post_2017_G_without_honors, paired = TRUE, p.adj = "bonf") cat("=============================\n") cat("Paired Student t-test of pre- vs post-recitation performance in 2017 Non-genie nohonors\n") t.test(score_pre_2017_NG_without_honors, score_post_2017_NG_without_honors, paired = TRUE, p.adj = "bonf") sink() ############################################################################################ #Visualization of performance on each year via violin plots VIOLINPLOT_2016_PRE_POST <- read.table("Sfile11_VIOLINPLOT_2016_PRE_POST.txt", header = TRUE, sep = "\t") head(VIOLINPLOT_2016_PRE_POST) VIOLINPLOT_2016_PRE_POST_without_honors <- VIOLINPLOT_2016_PRE_POST[!(VIOLINPLOT_2016_PRE_POST$class=="Pair1_730"),] head(VIOLINPLOT_2016_PRE_POST_without_honors) VIOLINPLOT_2017_GENIE_PRE_POST <- read.table("Sfile12_VIOLINPLOT_2017_GENIE_PRE_POST.txt", header = TRUE, sep = "\t") head(VIOLINPLOT_2017_GENIE_PRE_POST) VIOLINPLOT_2017_GENIE_PRE_POST_without_honors <- VIOLINPLOT_2017_GENIE_PRE_POST[!(VIOLINPLOT_2017_GENIE_PRE_POST$class=="Pair2_130"),] head(VIOLINPLOT_2017_GENIE_PRE_POST_without_honors) VIOLINPLOT_2017_NOGENIE_PRE_POST <- read.table("Sfile13_VIOLINPLOT_2017_NOGENIE_PRE_POST.txt", header = TRUE, sep = "\t") head(VIOLINPLOT_2017_NOGENIE_PRE_POST) require(gridExtra) VIOLINPLOT_2016_without_honors <- ggplot(VIOLINPLOT_2016_PRE_POST_without_honors, aes(x = year, y = score, fill=year)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete() + xlab("") + ylab("2016") + theme(legend.position = "none") + scale_fill_brewer(palette="Dark2") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") VIOLINPLOT_2016_without_honors <- VIOLINPLOT_2016_without_honors + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) VIOLINPLOT_2017_GENIE_without_honors <- ggplot(VIOLINPLOT_2017_GENIE_PRE_POST_without_honors, aes(x = year, y = score, fill=year)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete() + xlab("") + ylab("Evaluation scores 2017 (Genie)") + theme(legend.position = "none") + scale_fill_brewer(palette="Dark2") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") VIOLINPLOT_2017_GENIE_without_honors <- VIOLINPLOT_2017_GENIE_without_honors + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) VIOLINPLOT_2017_NOGENIE <- ggplot(VIOLINPLOT_2017_NOGENIE_PRE_POST, aes(x = year, y = score, fill=year)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete() + xlab("") + ylab("2017 (No Genie)") + theme(legend.position = "none") + scale_fill_brewer(palette="Dark2") + geom_violin(trim=FALSE) + geom_jitter(shape=16, position=position_jitter(0.1), col="black") VIOLINPLOT_2017_NOGENIE <- VIOLINPLOT_2017_NOGENIE + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) #Paired student plot pd <- position_dodge(0.01) PAIRED_STUDENT_2016_without_honors <- ggplot(VIOLINPLOT_2016_PRE_POST_without_honors, aes(x = year, y = score)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2016" = "Pre-recitation", "Post-recitation_score_2016" = "Post-recitation")) + xlab("") + labs(title="Genie 2016", y ="Scores")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + geom_line(aes(group = student, color=student), position = pd) + geom_point(position = pd) PAIRED_STUDENT_2017_GENIE_without_honors <- ggplot(VIOLINPLOT_2017_GENIE_PRE_POST_without_honors, aes(x = year, y = score)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2017" = "Pre-recitation", "Post-recitation_score_2017" = "Post-recitation")) + xlab("") + labs(title="Genie 2017", y ="")+ theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + geom_line(aes(group = student, color=student), position = pd) + geom_point(position = pd) PAIRED_STUDENT_2017_NOGENIE <- ggplot(VIOLINPLOT_2017_NOGENIE_PRE_POST, aes(x = year, y = score)) + scale_y_continuous(limits = c(0, 1.1), breaks = seq(0, 1.1, by = 0.1)) + scale_x_discrete(labels=c(".Pre-recitation_score_2017" = "Pre-recitation", "Post-recitation_score_2017" = "Post-recitation")) + xlab("") + labs(title="Non-Genie 2017", y ="") + theme_bw() + theme(panel.border = element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), axis.line = element_line(colour = "black")) + theme(legend.position = "none") + geom_line(aes(group = student, color=student), position = pd) + geom_point(position = pd) pdf(file="nohonors_Sfile14_violin_plots.pdf") FS14 <- ggarrange(PAIRED_STUDENT_2016_without_honors, PAIRED_STUDENT_2017_GENIE_without_honors, PAIRED_STUDENT_2017_NOGENIE, ncol=3) print(FS14) dev.off() ############################################################################################ #Performance by question compiled_years_by_question <- read.table("Sfile10_RESULTS_COMPILED_BY_QUESTION.txt", header = TRUE, sep = "\t") head(compiled_years_by_question) #The results for both years and treatments removing the honors recitation section on each year compiled_years_by_question_without_honors <- compiled_years_by_question[!(compiled_years_by_question$class=="Pair1_730" & compiled_years_by_question$year=="2016_Genie"),] compiled_years_by_question_without_honors <- compiled_years_by_question_without_honors[!(compiled_years_by_question_without_honors$class=="Pair2_130" & compiled_years_by_question_without_honors$year=="2017_Genie"),] head(compiled_years_by_question_without_honors) split_years_by_year <- split(compiled_years_by_question_without_honors, compiled_years_by_question_without_honors$year) df_2016_by_year_without_honors <- split_years_by_year[[1]] head(df_2016_by_year_without_honors) df_2017_G_by_year_without_honors <- split_years_by_year[[2]] head(df_2017_G_by_year_without_honors) df_2017_NG_by_year_without_honors <- split_years_by_year[[3]] head(df_2017_NG_by_year_without_honors) # Calculate McNemar's Test Statistic # https://en.wikipedia.org/wiki/McNemar%27s_test mcnemar_stat <- function(b,c) { (b-c)^2/(b+c) } # Reshape the input data so that each year-student-question has its own line. # Each observation will have a Pre and Post Stat dat <- compiled_years_by_question_without_honors %>% pivot_longer(!c(year,class,student)) %>% separate("name", c("quiz", "question"), sep="_") %>% mutate(question = as.integer(question)) %>% pivot_wider(names_from="quiz",values_from="value") # Count for each year-question the interaction of pre and post. # use these counts to measure McNemar's statistic and its pvalue. dat1 <- dat %>% mutate(result = glue("ans_{Pre}_{Post}")) %>% count(year, question, result) %>% pivot_wider(names_from="result",values_from="n", values_fill=list(n=0L)) %>% mutate(mcnemar_stat = mcnemar_stat(ans_1_0, ans_0_1), mcnemar_pvalue = pchisq(mcnemar_stat, 1L, lower.tail = FALSE)) # Save Results write_tsv(dat1,"nohonors_mcnemar.tsv") ############################################################################################ #Comparing 2017 genie vs non-genie classes performance per question rawdat <- read.table("Sfile10_RESULTS_COMPILED_BY_QUESTION.txt", header = TRUE, sep = "\t") head(rawdat) #The results for both years and treatments removing the honors recitation section on each year compiled_years_by_question_without_honors <- rawdat[!(rawdat$class=="Pair1_730" & rawdat$year=="2016_Genie"),] compiled_years_by_question_without_honors <- rawdat[!(rawdat$class=="Pair2_130" & rawdat$year=="2017_Genie"),] head(compiled_years_by_question_without_honors) dat0 <- compiled_years_by_question_without_honors %>% pivot_longer(!c(year,class,student)) %>% separate("name", c("quiz", "question"), sep="_") %>% mutate(question = as.integer(question)) %>% pivot_wider(names_from="quiz",values_from="value") write_tsv(dat0,"nohonors_pre_and_post.tsv") dat <- read_tsv("nohonors_pre_and_post.tsv") # Compare 2017 Genie and 2017 Not Genie on pre and post # Use Fisher's exact tests to test the association between question switching and # type of instruction. dat1 <- dat %>% mutate(result = str_glue("ans_{Pre}_{Post}")) %>% count(year, question, result) %>% pivot_wider(names_from="result",values_from="n", values_fill=list(n=0L)) %>% filter( year %in% c("2017_Genie", "2017_No_Genie")) dat2 <- dat1 %>% select(year,question,ans_0_1,ans_1_0) %>% pivot_wider(names_from="year", values_from=c(ans_0_1,ans_1_0)) f <- function(a,b,c,d) { m <- matrix(c(a,b,c,d),nrow=2) v <- fisher.test(m) data.frame(OR.est=v$estimate, p.value=v$p.value) } dat3 <- dat2 %>% rowwise() %>% summarize(f(ans_0_1_2017_Genie, ans_0_1_2017_No_Genie, ans_1_0_2017_Genie, ans_1_0_2017_No_Genie)) dat4 <- bind_cols(dat2,dat3) write_tsv(dat4,"nohonors_genie_vs_non_genie.tsv") ############################################################################################ pal_okabe_ito <- c("#E69F00", "#56B4E9", "#009E73", "#F0E442", "#0072B2", "#D55E00", "#CC79A7", "#000000") dat <- read_tsv("pre_and_post_nohonors.tsv") dat.A <- dat %>% mutate(result = str_glue("ans_{Pre}_{Post}")) %>% count(year, student, result) %>% pivot_wider(names_from="result",values_from="n", values_fill=list(n=0L)) %>% mutate(pre_acc = (ans_1_1+ans_1_0), post_acc = (ans_1_1+ans_0_1)) dat.B <- dat.A %>% mutate(diff = post_acc - pre_acc, diffs = sign(diff), excess = "no") %>% arrange(diff) dat4 <- dat.B %>% group_by(year) %>% group_modify(function(tbl,y){ j <- tail(1:nrow(tbl),sum(tbl$diffs)) tbl$excess[j] <- "yes" tbl }) pretty_year <- c( `2016_Genie` = "Genie 2016", `2017_Genie` = "Genie 2017", `2017_No_Genie` = "Non-Genie 2017" ) ann_tbl <- tibble(pre_acc=10,post_acc=2.5,year="2017_No_Genie",excess="yes") pdf(file="nohonors_F1.pdf") set.seed(0x90210) gg <- ggplot(dat4, aes(x=pre_acc,y=post_acc,color=excess)) + geom_jitter(width=0.3,height=0.3) gg <- gg + geom_abline(slope=1,intercept=0) gg <- gg + labs(x = "Pre-Score", y = "Post-Score") gg <- gg + theme_minimal() + coord_fixed(xlim=c(0,22),ylim=c(0,22)) gg <- gg + facet_wrap(~year, labeller = labeller(year = pretty_year)) gg <- gg + scale_color_manual(values=pal_okabe_ito[c(8,2)]) gg <- gg + theme(legend.position = "none") gg <- gg + geom_label(data=ann_tbl,label=" Excess\n Improvement",hjust="left", nudge_x=-2,color="black",size=3) + geom_point(data=ann_tbl,size=5) print(gg) dev.off() ############################################################################################ #Visualization per question #Transforms the pre-recitation scores columns on each data frame into a numeric vector score_pre1_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_1)) score_pre2_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_2)) score_pre3_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_3)) score_pre4_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_4)) score_pre5_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_5)) score_pre6_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_6)) score_pre7_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_7)) score_pre8_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_8)) score_pre9_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_9)) score_pre10_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_10)) score_pre11_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_11)) score_pre12_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_12)) score_pre13_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_13)) score_pre14_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_14)) score_pre15_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_15)) score_pre16_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_16)) score_pre17_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_17)) score_pre18_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_18)) score_pre19_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_19)) score_pre20_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_20)) score_pre21_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_21)) score_pre22_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Pre_22)) score_pre1_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_1)) score_pre2_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_2)) score_pre3_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_3)) score_pre4_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_4)) score_pre5_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_5)) score_pre6_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_6)) score_pre7_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_7)) score_pre8_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_8)) score_pre9_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_9)) score_pre10_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_10)) score_pre11_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_11)) score_pre12_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_12)) score_pre13_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_13)) score_pre14_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_14)) score_pre15_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_15)) score_pre16_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_16)) score_pre17_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_17)) score_pre18_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_18)) score_pre19_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_19)) score_pre20_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_20)) score_pre21_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_21)) score_pre22_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Pre_22)) #Gets score for that question score_pre1_2017_G_val_without_honors <- sum(score_pre1_2017_G_without_honors)/length(score_pre1_2017_G_without_honors) score_pre2_2017_G_val_without_honors <- sum(score_pre2_2017_G_without_honors)/length(score_pre2_2017_G_without_honors) score_pre3_2017_G_val_without_honors <- sum(score_pre3_2017_G_without_honors)/length(score_pre3_2017_G_without_honors) score_pre4_2017_G_val_without_honors <- sum(score_pre4_2017_G_without_honors)/length(score_pre4_2017_G_without_honors) score_pre5_2017_G_val_without_honors <- sum(score_pre5_2017_G_without_honors)/length(score_pre5_2017_G_without_honors) score_pre6_2017_G_val_without_honors <- sum(score_pre6_2017_G_without_honors)/length(score_pre6_2017_G_without_honors) score_pre7_2017_G_val_without_honors <- sum(score_pre7_2017_G_without_honors)/length(score_pre7_2017_G_without_honors) score_pre8_2017_G_val_without_honors <- sum(score_pre8_2017_G_without_honors)/length(score_pre8_2017_G_without_honors) score_pre9_2017_G_val_without_honors <- sum(score_pre9_2017_G_without_honors)/length(score_pre9_2017_G_without_honors) score_pre10_2017_G_val_without_honors <- sum(score_pre10_2017_G_without_honors)/length(score_pre10_2017_G_without_honors) score_pre11_2017_G_val_without_honors <- sum(score_pre11_2017_G_without_honors)/length(score_pre11_2017_G_without_honors) score_pre12_2017_G_val_without_honors <- sum(score_pre12_2017_G_without_honors)/length(score_pre12_2017_G_without_honors) score_pre13_2017_G_val_without_honors <- sum(score_pre13_2017_G_without_honors)/length(score_pre13_2017_G_without_honors) score_pre14_2017_G_val_without_honors <- sum(score_pre14_2017_G_without_honors)/length(score_pre14_2017_G_without_honors) score_pre15_2017_G_val_without_honors <- sum(score_pre15_2017_G_without_honors)/length(score_pre15_2017_G_without_honors) score_pre16_2017_G_val_without_honors <- sum(score_pre16_2017_G_without_honors)/length(score_pre16_2017_G_without_honors) score_pre17_2017_G_val_without_honors <- sum(score_pre17_2017_G_without_honors)/length(score_pre17_2017_G_without_honors) score_pre18_2017_G_val_without_honors <- sum(score_pre18_2017_G_without_honors)/length(score_pre18_2017_G_without_honors) score_pre19_2017_G_val_without_honors <- sum(score_pre19_2017_G_without_honors)/length(score_pre19_2017_G_without_honors) score_pre20_2017_G_val_without_honors <- sum(score_pre20_2017_G_without_honors)/length(score_pre20_2017_G_without_honors) score_pre21_2017_G_val_without_honors <- sum(score_pre21_2017_G_without_honors)/length(score_pre21_2017_G_without_honors) score_pre22_2017_G_val_without_honors <- sum(score_pre22_2017_G_without_honors)/length(score_pre22_2017_G_without_honors) score_pre1_2017_NG_val_without_honors <- sum(score_pre1_2017_NG_without_honors)/length(score_pre1_2017_NG_without_honors) score_pre2_2017_NG_val_without_honors <- sum(score_pre2_2017_NG_without_honors)/length(score_pre2_2017_NG_without_honors) score_pre3_2017_NG_val_without_honors <- sum(score_pre3_2017_NG_without_honors)/length(score_pre3_2017_NG_without_honors) score_pre4_2017_NG_val_without_honors <- sum(score_pre4_2017_NG_without_honors)/length(score_pre4_2017_NG_without_honors) score_pre5_2017_NG_val_without_honors <- sum(score_pre5_2017_NG_without_honors)/length(score_pre5_2017_NG_without_honors) score_pre6_2017_NG_val_without_honors <- sum(score_pre6_2017_NG_without_honors)/length(score_pre6_2017_NG_without_honors) score_pre7_2017_NG_val_without_honors <- sum(score_pre7_2017_NG_without_honors)/length(score_pre7_2017_NG_without_honors) score_pre8_2017_NG_val_without_honors <- sum(score_pre8_2017_NG_without_honors)/length(score_pre8_2017_NG_without_honors) score_pre9_2017_NG_val_without_honors <- sum(score_pre9_2017_NG_without_honors)/length(score_pre9_2017_NG_without_honors) score_pre10_2017_NG_val_without_honors <- sum(score_pre10_2017_NG_without_honors)/length(score_pre10_2017_NG_without_honors) score_pre11_2017_NG_val_without_honors <- sum(score_pre11_2017_NG_without_honors)/length(score_pre11_2017_NG_without_honors) score_pre12_2017_NG_val_without_honors <- sum(score_pre12_2017_NG_without_honors)/length(score_pre12_2017_NG_without_honors) score_pre13_2017_NG_val_without_honors <- sum(score_pre13_2017_NG_without_honors)/length(score_pre13_2017_NG_without_honors) score_pre14_2017_NG_val_without_honors <- sum(score_pre14_2017_NG_without_honors)/length(score_pre14_2017_NG_without_honors) score_pre15_2017_NG_val_without_honors <- sum(score_pre15_2017_NG_without_honors)/length(score_pre15_2017_NG_without_honors) score_pre16_2017_NG_val_without_honors <- sum(score_pre16_2017_NG_without_honors)/length(score_pre16_2017_NG_without_honors) score_pre17_2017_NG_val_without_honors <- sum(score_pre17_2017_NG_without_honors)/length(score_pre17_2017_NG_without_honors) score_pre18_2017_NG_val_without_honors <- sum(score_pre18_2017_NG_without_honors)/length(score_pre18_2017_NG_without_honors) score_pre19_2017_NG_val_without_honors <- sum(score_pre19_2017_NG_without_honors)/length(score_pre19_2017_NG_without_honors) score_pre20_2017_NG_val_without_honors <- sum(score_pre20_2017_NG_without_honors)/length(score_pre20_2017_NG_without_honors) score_pre21_2017_NG_val_without_honors <- sum(score_pre21_2017_NG_without_honors)/length(score_pre21_2017_NG_without_honors) score_pre22_2017_NG_val_without_honors <- sum(score_pre22_2017_NG_without_honors)/length(score_pre22_2017_NG_without_honors) #Transforms the post-recitation scores columns on each data frame into a numeric vector score_post1_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_1)) score_post2_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_2)) score_post3_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_3)) score_post4_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_4)) score_post5_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_5)) score_post6_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_6)) score_post7_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_7)) score_post8_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_8)) score_post9_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_9)) score_post10_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_10)) score_post11_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_11)) score_post12_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_12)) score_post13_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_13)) score_post14_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_14)) score_post15_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_15)) score_post16_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_16)) score_post17_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_17)) score_post18_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_18)) score_post19_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_19)) score_post20_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_20)) score_post21_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_21)) score_post22_2017_G_without_honors <- as.numeric(as.character(df_2017_G_by_year_without_honors$Post_22)) score_post1_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_1)) score_post2_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_2)) score_post3_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_3)) score_post4_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_4)) score_post5_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_5)) score_post6_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_6)) score_post7_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_7)) score_post8_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_8)) score_post9_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_9)) score_post10_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_10)) score_post11_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_11)) score_post12_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_12)) score_post13_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_13)) score_post14_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_14)) score_post15_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_15)) score_post16_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_16)) score_post17_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_17)) score_post18_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_18)) score_post19_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_19)) score_post20_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_20)) score_post21_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_21)) score_post22_2017_NG_without_honors <- as.numeric(as.character(df_2017_NG_by_year_without_honors$Post_22)) #Gets score for that question score_post1_2017_G_val_without_honors <- sum(score_post1_2017_G_without_honors)/length(score_post1_2017_G_without_honors) score_post2_2017_G_val_without_honors <- sum(score_post2_2017_G_without_honors)/length(score_post2_2017_G_without_honors) score_post3_2017_G_val_without_honors <- sum(score_post3_2017_G_without_honors)/length(score_post3_2017_G_without_honors) score_post4_2017_G_val_without_honors <- sum(score_post4_2017_G_without_honors)/length(score_post4_2017_G_without_honors) score_post5_2017_G_val_without_honors <- sum(score_post5_2017_G_without_honors)/length(score_post5_2017_G_without_honors) score_post6_2017_G_val_without_honors <- sum(score_post6_2017_G_without_honors)/length(score_post6_2017_G_without_honors) score_post7_2017_G_val_without_honors <- sum(score_post7_2017_G_without_honors)/length(score_post7_2017_G_without_honors) score_post8_2017_G_val_without_honors <- sum(score_post8_2017_G_without_honors)/length(score_post8_2017_G_without_honors) score_post9_2017_G_val_without_honors <- sum(score_post9_2017_G_without_honors)/length(score_post9_2017_G_without_honors) score_post10_2017_G_val_without_honors <- sum(score_post10_2017_G_without_honors)/length(score_post10_2017_G_without_honors) score_post11_2017_G_val_without_honors <- sum(score_post11_2017_G_without_honors)/length(score_post11_2017_G_without_honors) score_post12_2017_G_val_without_honors <- sum(score_post12_2017_G_without_honors)/length(score_post12_2017_G_without_honors) score_post13_2017_G_val_without_honors <- sum(score_post13_2017_G_without_honors)/length(score_post13_2017_G_without_honors) score_post14_2017_G_val_without_honors <- sum(score_post14_2017_G_without_honors)/length(score_post14_2017_G_without_honors) score_post15_2017_G_val_without_honors <- sum(score_post15_2017_G_without_honors)/length(score_post15_2017_G_without_honors) score_post16_2017_G_val_without_honors <- sum(score_post16_2017_G_without_honors)/length(score_post16_2017_G_without_honors) score_post17_2017_G_val_without_honors <- sum(score_post17_2017_G_without_honors)/length(score_post17_2017_G_without_honors) score_post18_2017_G_val_without_honors <- sum(score_post18_2017_G_without_honors)/length(score_post18_2017_G_without_honors) score_post19_2017_G_val_without_honors <- sum(score_post19_2017_G_without_honors)/length(score_post19_2017_G_without_honors) score_post20_2017_G_val_without_honors <- sum(score_post20_2017_G_without_honors)/length(score_post20_2017_G_without_honors) score_post21_2017_G_val_without_honors <- sum(score_post21_2017_G_without_honors)/length(score_post21_2017_G_without_honors) score_post22_2017_G_val_without_honors <- sum(score_post22_2017_G_without_honors)/length(score_post22_2017_G_without_honors) score_post1_2017_NG_val_without_honors <- sum(score_post1_2017_NG_without_honors)/length(score_post1_2017_NG_without_honors) score_post2_2017_NG_val_without_honors <- sum(score_post2_2017_NG_without_honors)/length(score_post2_2017_NG_without_honors) score_post3_2017_NG_val_without_honors <- sum(score_post3_2017_NG_without_honors)/length(score_post3_2017_NG_without_honors) score_post4_2017_NG_val_without_honors <- sum(score_post4_2017_NG_without_honors)/length(score_post4_2017_NG_without_honors) score_post5_2017_NG_val_without_honors <- sum(score_post5_2017_NG_without_honors)/length(score_post5_2017_NG_without_honors) score_post6_2017_NG_val_without_honors <- sum(score_post6_2017_NG_without_honors)/length(score_post6_2017_NG_without_honors) score_post7_2017_NG_val_without_honors <- sum(score_post7_2017_NG_without_honors)/length(score_post7_2017_NG_without_honors) score_post8_2017_NG_val_without_honors <- sum(score_post8_2017_NG_without_honors)/length(score_post8_2017_NG_without_honors) score_post9_2017_NG_val_without_honors <- sum(score_post9_2017_NG_without_honors)/length(score_post9_2017_NG_without_honors) score_post10_2017_NG_val_without_honors <- sum(score_post10_2017_NG_without_honors)/length(score_post10_2017_NG_without_honors) score_post11_2017_NG_val_without_honors <- sum(score_post11_2017_NG_without_honors)/length(score_post11_2017_NG_without_honors) score_post12_2017_NG_val_without_honors <- sum(score_post12_2017_NG_without_honors)/length(score_post12_2017_NG_without_honors) score_post13_2017_NG_val_without_honors <- sum(score_post13_2017_NG_without_honors)/length(score_post13_2017_NG_without_honors) score_post14_2017_NG_val_without_honors <- sum(score_post14_2017_NG_without_honors)/length(score_post14_2017_NG_without_honors) score_post15_2017_NG_val_without_honors <- sum(score_post15_2017_NG_without_honors)/length(score_post15_2017_NG_without_honors) score_post16_2017_NG_val_without_honors <- sum(score_post16_2017_NG_without_honors)/length(score_post16_2017_NG_without_honors) score_post17_2017_NG_val_without_honors <- sum(score_post17_2017_NG_without_honors)/length(score_post17_2017_NG_without_honors) score_post18_2017_NG_val_without_honors <- sum(score_post18_2017_NG_without_honors)/length(score_post18_2017_NG_without_honors) score_post19_2017_NG_val_without_honors <- sum(score_post19_2017_NG_without_honors)/length(score_post19_2017_NG_without_honors) score_post20_2017_NG_val_without_honors <- sum(score_post20_2017_NG_without_honors)/length(score_post20_2017_NG_without_honors) score_post21_2017_NG_val_without_honors <- sum(score_post21_2017_NG_without_honors)/length(score_post21_2017_NG_without_honors) score_post22_2017_NG_val_without_honors <- sum(score_post22_2017_NG_without_honors)/length(score_post22_2017_NG_without_honors) #Preparing vectors for the dataframe #Order based on Figure2a -Key concepts| Table 3 - Misconceptions questionA2T3 <- c("Q1","Q3","Q15","Q4","Q10","Q13","Q16","Q7","Q5","Q6","Q8","Q2","Q9","Q12","Q17","Q20","Q14","Q19","Q22","Q11","Q18","Q21") time_pre <- rep("Pre", 22) time_post <- rep("Post", 22) #Vector of pre-recitation and post-recitation scores by question on 2017 genie classes pre_recitation_scores_2017_G_without_honorsA2T3 <- c(score_pre1_2017_G_val_without_honors, score_pre3_2017_G_val_without_honors, score_pre15_2017_G_val_without_honors, score_pre4_2017_G_val_without_honors, score_pre10_2017_G_val_without_honors, score_pre13_2017_G_val_without_honors, score_pre16_2017_G_val_without_honors, score_pre7_2017_G_val_without_honors, score_pre5_2017_G_val_without_honors, score_pre6_2017_G_val_without_honors, score_pre8_2017_G_val_without_honors, score_pre2_2017_G_val_without_honors, score_pre9_2017_G_val_without_honors, score_pre12_2017_G_val_without_honors, score_pre17_2017_G_val_without_honors, score_pre20_2017_G_val_without_honors, score_pre14_2017_G_val_without_honors, score_pre19_2017_G_val_without_honors, score_pre22_2017_G_val_without_honors, score_pre11_2017_G_val_without_honors, score_pre18_2017_G_val_without_honors, score_pre21_2017_G_val_without_honors) post_recitation_scores_2017_G_without_honorsA2T3 <- c(score_post1_2017_G_val_without_honors, score_post3_2017_G_val_without_honors, score_post15_2017_G_val_without_honors, score_post4_2017_G_val_without_honors, score_post10_2017_G_val_without_honors, score_post13_2017_G_val_without_honors, score_post16_2017_G_val_without_honors, score_post7_2017_G_val_without_honors, score_post5_2017_G_val_without_honors, score_post6_2017_G_val_without_honors, score_post8_2017_G_val_without_honors, score_post2_2017_G_val_without_honors, score_post9_2017_G_val_without_honors, score_post12_2017_G_val_without_honors, score_post17_2017_G_val_without_honors, score_post20_2017_G_val_without_honors, score_post14_2017_G_val_without_honors, score_post19_2017_G_val_without_honors, score_post22_2017_G_val_without_honors, score_post11_2017_G_val_without_honors, score_post18_2017_G_val_without_honors, score_post21_2017_G_val_without_honors) #Pre, post-, and combined recitations dataframes for plot 2017 genie class df_scores_by_question_pre_2017_G_without_honorsA2T3 <- data.frame(scores_2017_G_without_honors = pre_recitation_scores_2017_G_without_honorsA2T3, time_of_recitation = time_pre, question=questionA2T3) df_scores_by_question_post_2017_G_without_honorsA2T3 <- data.frame(scores_2017_G_without_honors = post_recitation_scores_2017_G_without_honorsA2T3, time_of_recitation = time_post, question=questionA2T3) df_scores_by_question_combined_2017_G_without_honorsA2T3 <- rbind(df_scores_by_question_pre_2017_G_without_honorsA2T3, df_scores_by_question_post_2017_G_without_honorsA2T3) #Vector of pre-recitation and post-recitation scores by question on 2017 no genie classes pre_recitation_scores_2017_NGA2T3 <- c(score_pre1_2017_NG_val_without_honors, score_pre3_2017_NG_val_without_honors, score_pre15_2017_NG_val_without_honors, score_pre4_2017_NG_val_without_honors, score_pre10_2017_NG_val_without_honors, score_pre13_2017_NG_val_without_honors, score_pre16_2017_NG_val_without_honors, score_pre7_2017_NG_val_without_honors, score_pre5_2017_NG_val_without_honors, score_pre6_2017_NG_val_without_honors, score_pre8_2017_NG_val_without_honors, score_pre2_2017_NG_val_without_honors, score_pre9_2017_NG_val_without_honors, score_pre12_2017_NG_val_without_honors, score_pre17_2017_NG_val_without_honors, score_pre20_2017_NG_val_without_honors, score_pre14_2017_NG_val_without_honors, score_pre19_2017_NG_val_without_honors, score_pre22_2017_NG_val_without_honors, score_pre11_2017_NG_val_without_honors, score_pre18_2017_NG_val_without_honors, score_pre21_2017_NG_val_without_honors) post_recitation_scores_2017_NGA2T3 <- c(score_post1_2017_NG_val_without_honors, score_post3_2017_NG_val_without_honors, score_post15_2017_NG_val_without_honors, score_post4_2017_NG_val_without_honors, score_post10_2017_NG_val_without_honors, score_post13_2017_NG_val_without_honors, score_post16_2017_NG_val_without_honors, score_post7_2017_NG_val_without_honors, score_post5_2017_NG_val_without_honors, score_post6_2017_NG_val_without_honors, score_post8_2017_NG_val_without_honors, score_post2_2017_NG_val_without_honors, score_post9_2017_NG_val_without_honors, score_post12_2017_NG_val_without_honors, score_post17_2017_NG_val_without_honors, score_post20_2017_NG_val_without_honors, score_post14_2017_NG_val_without_honors, score_post19_2017_NG_val_without_honors, score_post22_2017_NG_val_without_honors, score_post11_2017_NG_val_without_honors, score_post18_2017_NG_val_without_honors, score_post21_2017_NG_val_without_honors) #Pre, post-, and combined recitations dataframes for plot 2017 no genie class df_scores_by_question_pre_2017_NGA2T3 <- data.frame(scores_2017_NG = pre_recitation_scores_2017_NGA2T3, time_of_recitation = time_pre, question=questionA2T3) df_scores_by_question_post_2017_NGA2T3 <- data.frame(scores_2017_NG = post_recitation_scores_2017_NGA2T3, time_of_recitation = time_post, question=questionA2T3) df_scores_by_question_combined_2017_NGA2T3 <- rbind(df_scores_by_question_pre_2017_NGA2T3, df_scores_by_question_post_2017_NGA2T3) #Diff without honors score_diff1_2017_G_without_honors <- score_post1_2017_G_val_without_honors-score_pre1_2017_G_val_without_honors score_diff2_2017_G_without_honors <- score_post2_2017_G_val_without_honors-score_pre2_2017_G_val_without_honors score_diff3_2017_G_without_honors <- score_post3_2017_G_val_without_honors-score_pre3_2017_G_val_without_honors score_diff4_2017_G_without_honors <- score_post4_2017_G_val_without_honors-score_pre4_2017_G_val_without_honors score_diff5_2017_G_without_honors <- score_post5_2017_G_val_without_honors-score_pre5_2017_G_val_without_honors score_diff6_2017_G_without_honors <- score_post6_2017_G_val_without_honors-score_pre6_2017_G_val_without_honors score_diff7_2017_G_without_honors <- score_post7_2017_G_val_without_honors-score_pre7_2017_G_val_without_honors score_diff8_2017_G_without_honors <- score_post8_2017_G_val_without_honors-score_pre8_2017_G_val_without_honors score_diff9_2017_G_without_honors <- score_post9_2017_G_val_without_honors-score_pre9_2017_G_val_without_honors score_diff10_2017_G_without_honors <- score_post10_2017_G_val_without_honors-score_pre10_2017_G_val_without_honors score_diff11_2017_G_without_honors <- score_post11_2017_G_val_without_honors-score_pre11_2017_G_val_without_honors score_diff12_2017_G_without_honors <- score_post12_2017_G_val_without_honors-score_pre12_2017_G_val_without_honors score_diff13_2017_G_without_honors <- score_post13_2017_G_val_without_honors-score_pre13_2017_G_val_without_honors score_diff14_2017_G_without_honors <- score_post14_2017_G_val_without_honors-score_pre14_2017_G_val_without_honors score_diff15_2017_G_without_honors <- score_post15_2017_G_val_without_honors-score_pre15_2017_G_val_without_honors score_diff16_2017_G_without_honors <- score_post16_2017_G_val_without_honors-score_pre16_2017_G_val_without_honors score_diff17_2017_G_without_honors <- score_post17_2017_G_val_without_honors-score_pre17_2017_G_val_without_honors score_diff18_2017_G_without_honors <- score_post18_2017_G_val_without_honors-score_pre18_2017_G_val_without_honors score_diff19_2017_G_without_honors <- score_post19_2017_G_val_without_honors-score_pre19_2017_G_val_without_honors score_diff20_2017_G_without_honors <- score_post20_2017_G_val_without_honors-score_pre20_2017_G_val_without_honors score_diff21_2017_G_without_honors <- score_post21_2017_G_val_without_honors-score_pre21_2017_G_val_without_honors score_diff22_2017_G_without_honors <- score_post22_2017_G_val_without_honors-score_pre22_2017_G_val_without_honors diff_recitation_scores_2017_G_without_honors2AT3 <- c(score_diff1_2017_G_without_honors, score_diff3_2017_G_without_honors, score_diff4_2017_G_without_honors, score_diff10_2017_G_without_honors, score_diff13_2017_G_without_honors, score_diff15_2017_G_without_honors, score_diff16_2017_G_without_honors, score_diff7_2017_G_without_honors, score_diff5_2017_G_without_honors, score_diff6_2017_G_without_honors, score_diff8_2017_G_without_honors, score_diff2_2017_G_without_honors, score_diff9_2017_G_without_honors, score_diff12_2017_G_without_honors, score_diff17_2017_G_without_honors, score_diff20_2017_G_without_honors, score_diff14_2017_G_without_honors, score_diff19_2017_G_without_honors, score_diff22_2017_G_without_honors, score_diff11_2017_G_without_honors, score_diff18_2017_G_without_honors, score_diff21_2017_G_without_honors) diff_recitation_scores_2017_G_without_honorsA2 <- c(score_diff1_2017_G_without_honors, score_diff3_2017_G_without_honors, score_diff4_2017_G_without_honors, score_diff10_2017_G_without_honors, score_diff13_2017_G_without_honors, score_diff15_2017_G_without_honors, score_diff16_2017_G_without_honors, score_diff21_2017_G_without_honors, score_diff22_2017_G_without_honors, score_diff19_2017_G_without_honors, score_diff7_2017_G_without_honors, score_diff9_2017_G_without_honors, score_diff20_2017_G_without_honors, score_diff18_2017_G_without_honors, score_diff11_2017_G_without_honors, score_diff2_2017_G_without_honors, score_diff14_2017_G_without_honors, score_diff5_2017_G_without_honors, score_diff17_2017_G_without_honors, score_diff12_2017_G_without_honors, score_diff6_2017_G_without_honors, score_diff8_2017_G_without_honors) ###Order Fig2|Table3 in Price et al 2014 time_diffG <- rep("Diff-score Genie", 22) time_diffNG <- rep("Diff-score no Genie", 22) df_diff_scores_by_question_G_without_honors2AT3 <- data.frame(diff_recitation_scores = diff_recitation_scores_2017_G_without_honors2AT3, question=question2AT3, time=time_diffG) df_diff_scores_by_question_NG2AT3 <- data.frame(diff_recitation_scores = diff_recitation_scores_2017_NG2AT3, question=question2AT3, time=time_diffNG) df_diff_scores_by_question_G_without_honors2AT3$question <- factor(df_diff_scores_by_question_G_without_honors2AT3$question, levels = df_diff_scores_by_question_G_without_honors2AT3$question) df_diff_scores_by_question_NG2AT3$question <- factor(df_diff_scores_by_question_NG2AT3$question, levels = df_diff_scores_by_question_NG2AT3$question) df_diff_scores_by_question_combined2AT3 <- rbind(df_diff_scores_by_question_G_without_honors2AT3, df_diff_scores_by_question_NG2AT3) pdf(file="nohonors_F3.pdf") my_purple_colors <- c("orchid4","plum") diff_per_question_plot2AT3 <- ggplot(data=df_diff_scores_by_question_combined2AT3, aes(x=question, y=diff_recitation_scores, fill=time)) + geom_bar(width=0.4, stat="identity", position=position_dodge(0.5)) + scale_fill_manual(values = my_purple_colors) + xlab("Question") + ylab("Scores") + labs(fill = "Recitation") + scale_x_discrete(limits= df_diff_scores_by_question_combined2AT3$p.question) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) print(diff_per_question_plot2AT3) dev.off() time_postG <- rep("Post-score Genie", 22) time_postNG <- rep("Post-score no Genie", 22) df_post_scores_by_question_G_without_honors2AT3 <- data.frame(post_recitation_scores = post_recitation_scores_2017_G_without_honorsA2T3, question=question2AT3, time=time_postG) df_post_scores_by_question_NG2AT3 <- data.frame(post_recitation_scores = post_recitation_scores_2017_NG2AT3, question=question2AT3, time=time_postNG) df_post_scores_by_question_G_without_honors2AT3$question <- factor(df_post_scores_by_question_G_without_honors2AT3$question, levels = df_post_scores_by_question_G_without_honors2AT3$question) df_post_scores_by_question_NG2AT3$question <- factor(df_post_scores_by_question_NG2AT3$question, levels = df_post_scores_by_question_NG2AT3$question) df_post_scores_by_question_combined2AT3 <- rbind(df_post_scores_by_question_G_without_honors2AT3, df_post_scores_by_question_NG2AT3) pdf(file="nohonors_F2.pdf") my_green_colors <- c("darkgreen","darkseagreen3") post_per_question_plot2AT3 <- ggplot(data=df_post_scores_by_question_combined2AT3, aes(x=question, y=post_recitation_scores, fill=time)) + geom_bar(width=0.4, stat="identity", position=position_dodge(0.5)) + scale_fill_manual(values = my_green_colors) + xlab("Question") + ylab("Scores") + labs(fill = "Recitation") + scale_x_discrete(limits= df_post_scores_by_question_combined2AT3$p.question) + theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) print(post_per_question_plot2AT3) dev.off()