wideToLong: Reshape from wide to long
In lsr: Companion to "Learning Statistics with R"

Description Usage Arguments Details Value See Also Examples

Reshape a data frame from wide form to long form using the variable names

1	wideToLong(data, within = "within", sep = "_", split = TRUE)

`data`	The data frame.
`within`	Name to give to the long-form within-subject factor(s)
`sep`	Separator string used in wide-form variable names
`split`	Should multiple within-subject factors be split into multiple variables?

The wideToLong function is the companion function to longToWide. The data argument is a "wide form" data frame, in which each row corresponds to a single experimental unit (e.g., a single subject). The output is a "long form" data frame, in which each row corresponds to a single observation.

The wideToLong function relies on the variable names to determine how the data should be reshaped. The naming scheme for these variables places the name of the measured variable first, followed by the levels of the within-subjects variable(s), separated by the separator string sep (default is _) The separator string cannot appear anywhere else in the variable names: variables without the separator string are assumed to be between-subject variables.

If the experiment measured the accuracy of participants at some task at two different points in time, then the wide form data frame would contain variables of the form accuracy_t1 and accuracy_t2. After reshaping, the long form data frame would contain one measured variable called accuracy, and a within-subjects factor with levels t1 and t2. The name of the within-subjects factor is the within argument.

The function supports experimental designs with multiple within-subjects factors and multi-variable observations. For example, suppose each experimental subject is tested in two conditions (cond1 and cond2), on each of two days (day1 and day2), yielding an experimental design in which four observations are made for each subject. For each such observation, we record the mean response time MRT for and proportion of correct responses PC for the participant. The variable names needed for a design such as this one would be MRT_cond1_day1, MRT_cond1_day2, PC_cond1_day1, etc. The within argument should be a vector of names for the within-subject factors: in this case, within = c("condition","day").

By default, if there are multiple within-subject factors implied by the existence of multiple separators, the output will keep these as distinct variables in the long form data frame (split=FALSE). If split=TRUE, the within-subject factors will be collapsed into a single variable.

A data frame containing the reshaped data

longToWide, reshape

# Outcome measure is mean response time (MRT), measured in two conditions
# with 4 participants. All participants participate in both conditions.

wide <- data.frame( accuracy_t1 = c( .15,.50,.78,.55 ),  # accuracy at time point 1
                    accuracy_t2 = c( .55,.32,.99,.60 ),  # accuracy at time point 2
                    id = 1:4 )                           # id variable

# convert to long form
wideToLong( wide, "time" )


# A more complex design with multiple within-subject factors. Again, we have only
# four participants, but now we have two different outcome measures, mean response
# time (MRT) and the proportion of correct responses (PC). Additionally, we have two
# different repeated measures variables. As before, we have the experimental condition
# (cond1, cond2), but this time each participant does both conditions on two different
# days (day1, day2). Finally, we have multiple between-subject variables too, namely
# id and gender.

wide2 <- data.frame( id = 1:4,
                     gender = factor( c("male","male","female","female") ),
                     MRT_cond1_day1 = c( 415,500,478,550 ),
                     MRT_cond2_day1 = c( 455,532,499,602 ),
                     MRT_cond1_day2 = c( 400,490,468,502 ),
                     MRT_cond2_day2 = c( 450,518,474,588 ),
                     PC_cond1_day1 = c( 79,83,91,75 ),
                     PC_cond2_day1 = c( 82,86,90,78 ),
                     PC_cond1_day2 = c( 88,92,98,89 ),
                     PC_cond2_day2 = c( 93,97,100,95 ) )

# conversion to long form:
wideToLong( wide2 )
wideToLong( wide2, within = c("condition","day") )

# treat "condition x day" as a single repeated measures variable:
wideToLong( wide2, split = FALSE)