210608_DoE_Pilots2021_TechChech.Rmd

---
title: "CAL ScratchJr TechCheck prepost"
output: html_notebook
---

# Preliminaries

## About the data

Data are from pilots of the CAL ScratchJr curriculum in the spring of 2021. 

## Read in the data

```{r}
techcheck <- read.csv("20210609_DOEpilot_AllKthru2_TechCheckDataAnon.csv", strip.white = T, stringsAsFactors = T, na.strings = c("", NA))
```

## Loading libraries

```{r}
checkpoint("2021-06-10")
```

## Color palettes

```{r colour_palettes, include=FALSE}
cbPalette <- c("#999999", "#E69F00", "#56B4E9", "#009E73", "#F0E442", "#0072B2", "#D55E00", "#CC79A7")
cbbPalette <- c("#999999","#E69F00", "#56B4E9", "#009E73", "#F0E442", "#0072B2", "#D55E00", "#CC79A7","#5C5C5C")
```

## Functions

### Min, max, mean, SD

```{r}
min_max_mean_sd <- function(variable){
  
  min = round(min(variable, na.rm = T),1)
  max = round(max(variable, na.rm = T),1)
  mean = round(mean(variable, na.rm = T),1)
  sd = round(sd(variable, na.rm = T),1)
  
  my_list = list(min, max, mean, sd) 
  
  return(my_list)
}
```

## Vectors

# Data prep

## Selecting variables

```{r}
techcheck <- techcheck %>% dplyr::select(Grade, TeacherID = Teacher_ID, StudentID = Student_ID, Gender, SchoolID = School_ID, TimeOfTest = TestTimePoint, TotalScore = totalscore)
```

## Modify & create variables

```{r}
techcheck$Grade <- recode(techcheck$Grade,  "Second Grade" = "SecondGrade", "First Grade"= "FirstGrade", "Kindergarten" = "Kindergarten")

techcheck$Grade <- factor(techcheck$Grade, levels = c("Kindergarten", "FirstGrade", "SecondGrade"))
```

```{r}
techcheck$StudentID <- as.factor(as.character(techcheck$StudentID))
```

```{r}
techcheck$TimeOfTest <- recode (techcheck$TimeOfTest, "Post- Curriculum" = "Post-Curriculum")
techcheck$TimeOfTest <- factor(techcheck$TimeOfTest, levels = c("Pre-Curriculum", "Post-Curriculum"))
```

```{r}
techcheck_no_na <- na.omit(techcheck)
```



## Make wide df

```{r}
techcheck_pre <- subset(techcheck, TimeOfTest=="Pre-Curriculum")
techcheck_pre <- techcheck_pre %>% dplyr::rename(PreCurriculum = TotalScore) %>% dplyr::select(-TimeOfTest)
techcheck_post <- techcheck %>% filter(TimeOfTest == "Post-Curriculum") %>% dplyr::select(StudentID, TimeOfTest, TotalScore) 
techcheck_post <- techcheck_post %>% dplyr::select(PostCurriculum = TotalScore, StudentID)
techcheck_wide <- right_join(techcheck_pre, techcheck_post, by = "StudentID")
```

```{r}
techcheck_wide$ChangeScore <- techcheck_wide$PostCurriculum - techcheck_wide$PreCurriculum
```

```{r}
techcheck_wide_no_na <- na.omit(techcheck_wide)
```


# Visualization

## Raincloud

```{r}

w = 6
h = 4

y_lim_min = 0
y_lim_max = 15

before = techcheck_wide$PreCurriculum
after = techcheck_wide$PostCurriculum
n <- length(before) 
d <- data.frame(y = c(before, after),
               x = rep(c(1,2), each=n),
               id = techcheck_wide$StudentID)
set.seed(321)
d$xj <- jitter(d$x, amount = .09) 
```

```{r}
techcheck_prepost_raincloud <- ggplot(data = d, aes(y = y)) +
   geom_point(data = d %>% filter(x =="1"), aes(x = xj), color = '#999999', size = 1.5, 
              alpha = .6) +
   geom_point(data = d %>% filter(x =="2"), aes(x = xj), color = '#E69F00', size = 1.5, 
              alpha = .6) +
   geom_line(aes(x = xj, group = id), color = 'lightgray', alpha = .3) +
  
   geom_half_boxplot(
     data = d %>% filter(x=="1"), aes(x=x, y = y), position = position_nudge(x = -.25), 
     side = "r",outlier.shape = NA, center = TRUE, errorbar.draw = FALSE, width = .2, 
     fill = '#999999') +
   
   geom_half_boxplot(
     data = d %>% filter(x=="2"), aes(x=x, y = y), position = position_nudge(x = .15), 
     side = "r",outlier.shape = NA, center = TRUE, errorbar.draw = FALSE, width = .2, 
     fill = '#E69F00') +
  
   geom_half_violin(
     data = d %>% filter(x=="1"),aes(x = x, y = y), position = position_nudge(x = -.3), 
     side = "l", fill = '#999999') +
 
   geom_half_violin(
     data = d %>% filter(x=="2"),aes(x = x, y = y), position = position_nudge(x = .3), 
     side = "r", fill = "#E69F00") +
  
  scale_x_continuous(breaks=c(1,2), labels=c("Pre", "Post"), limits=c(0, 3)) +
   xlab("Time point") + ylab("TechCheck score") +
   theme_classic()+
   coord_cartesian(ylim=c(y_lim_min, y_lim_max))

ggsave("210609_Pilots2021_TechCheck_prepost.png")
```
## Barchart 

Create summary

```{r}
techcheck_summary <- techcheck %>% 
  group_by(Grade, TimeOfTest) %>% 
  summarise(mean_score = mean(TotalScore),  
            sd_score = sd(TotalScore), 
            n_score = n(),
            SE_score = sd(TotalScore)/sqrt(n()))

techcheck_summary <- na.omit(techcheck_summary)

techcheck_summary$Grade <- factor(techcheck_summary$Grade, levels = c("Kindergarten", "FirstGrade", "SecondGrade"))
techcheck_barplot <- ggplot(techcheck_summary, aes(x = TimeOfTest, y = mean_score, Group = Grade, fill = TimeOfTest)) +
  geom_bar(stat = "summary", fun.y = "mean")+
  geom_errorbar(aes(ymin = mean_score-sd_score, ymax = mean_score+sd_score,width=0.2))+
  scale_fill_manual(values = cbbPalette)+
  theme_minimal()+
  facet_wrap(~Grade)+
  theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust = 1))
ggsave("210609_Pilots2021_TechCheck_prepost_by_grade.png")
```
```{r}
techcheck_boxplot <- ggplot(aes(y = TotalScore, x = Grade, fill= TimeOfTest), data = techcheck_no_na) +
  geom_boxplot()+
  scale_fill_manual(values = cbbPalette)+
  theme_minimal()+
  theme(text = element_text(size = 20))   
ggsave("210609_TechCheck_boxplot.png")
```

## Gender difs

```{r}
techcheck_gender_boxplot <- ggplot(aes(y = TotalScore, x = Gender, fill = Gender), data = techcheck_no_na) +
  geom_boxplot()+
  theme_minimal()+
  theme(text = element_text(size = 20))+
  ylab ("TechCheck TotalScore")
ggsave("210609_TechCheck_gender_boxplot.png")
```


# Analysis

## Descriptives

```{r}
min_max_mean_sd(techcheck$TotalScore[techcheck$Grade=="Kindergarten"])
min_max_mean_sd(techcheck$TotalScore[techcheck$Grade=="FirstGrade"])
min_max_mean_sd(techcheck$TotalScore[techcheck$Grade=="SecondGrade"])
```

```{r}
min_max_mean_sd(techcheck_wide$ChangeScore[techcheck_wide$Grade=="Kindergarten"])
min_max_mean_sd(techcheck_wide$ChangeScore[techcheck_wide$Grade=="FirstGrade"])
min_max_mean_sd(techcheck_wide$ChangeScore[techcheck_wide$Grade=="SecondGrade"])
```