# HW1 (Due Sep 13) Key

# Change working directory to wherever your files are stored

# 1. Read in data, inspect, DVs and IVs
d = dfReadDat('HW1_Data.dat')                       ##Read in data file
str(d)                                             ##View variables
# IV/predictor: Condition, PreTest
# DV: PostCon, PostPerc, PostProc


# 2.
# a. Summary
summary(d)

# b. Descriptives
varDescribe(d)

# c. Just for pretest
varDescribe(d$PreTest)

# d. Histogram for pretest
hist(d$PreTest, main = 'PreTest Scores')

# e. diff for p. 9
d[9,'PostCon'] - d[9,'PostPerc']


# 3. mean-center pre-test
d$PreTestC = d$PreTest - mean(d$PreTest)
# check:
mean(d$PreTestC)


# 4. center condition
d$ConditionC = varRecode(d$Condition,c(0,1), c(-.5, .5))
mean(d$ConditionC)


# 5. global learning measure
# a. 
d$PostTestM = rowMeans(cbind(d$PostCon, d$PostProc, d$PostPerc), na.rm=TRUE)
# This is saying: combining the variables PostCon, PostProc, PostPerc into a new matrix (with only three 
# columns, each corrersponding to one variable) calculate the means of each row for this new matrix and 
# enter the value into this new variable named 'PostTestT', remove (ignore) missing values

# b.
varDescribe(d$PostTestM) # The mean is 73.93 and the sd is 9.74

# c.
varDescribeBy(d$PostTestM, d$ConditionC)

# 6. standardize postcon
# a.
d$PostConZ = (d$PostCon - mean(d$PostCon)) / sd(d$PostCon)

# b.
varDescribe(d$PostConZ) ##mean=0, sd=1

# c.
str(d)

#d.
varDescribe(d)


# 7. table
varDescribe(d) ## See word key for the table

# 8. plotting
plot(d$PreTest, d$PostTestM, main='Mean PostTest scores as function of PreTest scores', xlab='PreTest', ylab='PostTest (composite)')
abline(lm(d$PostTestM ~ d$PreTest))

# 9. outcome by condition
varDescribeBy(d$PostTestM, d$Condition)