#### Analysis for Study 1: Introductory Chemistry ####

setwd(" ") 


#### Import data ####

data = read.csv("Study1_data.csv")


#### Identify performance gaps (Figure 1) ####

# separate control section
data_control = data %>%
  filter(intervention == "No") 

# identify effects of the 4 demographic variables, with standard errors
control_gaps = lm(total_percent ~ gender_num_centered + fgen_num_centered + 
                                 urm_num_centered + asian_num_centered + act_centered +
                                 gpa_centered, data = data_control, na.action = "na.omit")
summary(control_gaps)

# URM students: B = -6.03905, SE = 2.62335, t(234) = -2.302, p = 0.02221


#### Analyze main effect of the intervention ####

model = lm(total_percent ~ intervention_num + act_centered + gpa_centered + 
                          gender_num_centered + fgen_num_centered + urm_num_centered +
                          asian_num_centered, data = data, na.action = "na.omit")
summary(model)

# Intervention: B = 0.6102, SE = 1.0916, t(527) = 0.559, p = 0.5764


#### Analyze interaction between URM status and the intervention (Figure 2) ####

model = lm(total_percent ~ intervention_num*urm_num_centered +
                            act_centered + gpa_centered + gender_num_centered +
                            asian_num_centered + fgen_num_centered, 
                            data = data, na.action = "na.omit")
summary(model)

# Intervention*URM: p = 0.0298


#### Find mean values and standard errors for the effect of the intervention on non-URM students

model = lm(total_percent ~ intervention_num*urm_num +
             act_centered + gpa_centered + gender_num_centered + 
             asian_num_centered + fgen_num_centered, data = data, 
           na.action = "na.omit")
summary(model)

# Intervention: B = -0.1892, SE = 1.1480, t(526) = -0.165, p = 0.8692


#### Find mean values and standard errors for the effect of the intervention on URM students

model = lm(total_percent ~ intervention_num*nonurm_num +
             act_centered + gpa_centered + gender_num_centered + 
             asian_num_centered + fgen_num_centered, data = data, 
           na.action = "na.omit")
summary(model)

# Intervention: B = 7.5410, SE = 3.3628, t(526) = 2.242, p = 0.0253