Merge ILSA files

In ILSAmerge: Merge and Download International Large-Scale Assessments (ILSA) Data

knitr::opts_chunk$set(
  collapse = TRUE
)

library(ILSAmerge)

Downloaded data

As an example, we can use the data of 'REDS' 2021 provided by the package.

In the case of 'REDS' 2021, we have 12 files and 3 populations:

ILSAfile.info(inputdir = system.file("extdata/reds", package = "ILSAmerge"))

With ILSAmerge() we can combine all the files into the number of populations we have.

The default type will be an "rds", but also we can make "zsav" and "sav" files:

ILSAmerge(inputdir = system.file("extdata/reds", package = "ILSAmerge"), 
          outputdir = tempdir(), 
          filetype = c("rds", "zsav", "sav"))

Adjust merge options

Depending on your goals or the capabilities of your computer, you may consider adjusting some options of ILSAmerge():

• population: If you need to merge only some of the populations from the path, you could include them here as a character vector. It should be in the same format as the output of ILSAfile.info().
• MBlimit: If set to NULL, the default, all files will be merged within R. If set a numeric value that will establish a limit (from the output of ILSAfile.info()) of which files are going to be merged by R, the populations that go over this limit will not be merged by R, but instead an SPSS syntax will be produced. Later, you can use this syntax to merge the files.
• MBlistlimit: In ILSAmerge() files are merged in two ways: using a list produced by an lapply function, or an empty matrix filled up by subscripts. The first method is almost always faster, but it can use a lot of memory. This argument establishes a limit for merging by list and not by matrix. We recommend not setting this higher than 200.
• SPSSlimit: If SPSS syntaxes are going to be produced it is important to take into account the SPSS requirements for handling files. This could vary by SPSS version, but normally it is a total of 50 files per command.

Progress and time for merging

When ILSAmerge() runs, information about the progress of the merging and the running time will be produced:

ILSAmerge(inputdir = system.file("extdata/reds", package = "ILSAmerge"), 
          outputdir = tempdir(), 
          filetype = c("rds", "zsav", "sav"))

Merging between respondents

After merging countries it is also possible to merge datasets of the same or different respondents.

Two functions are available to merge respondents: combineStudents(), and addSchools.

Combine students

combineStudents() combines multiple student files, e.g., achievement and background questionnaires, using 'TIMSS' Advanced 1995:

unlink(file.path(tempdir(),"combineStudents"),recursive = TRUE)

# Path were raw 'SPSS' files are
input <- system.file("extdata/timssadv", package = "ILSAmerge")

# Path were merged files will be saved
dir.create(file.path(tempdir(),"combineStudents"))
output <- file.path(tempdir(),"combineStudents")

# Merging 'TIMSS' Advanced 1995, as .rds file
ILSAmerge(inputdir = input, outputdir = output, filetype = "rds", quiet = TRUE)

# Rename files
ILSArename(output)

# Check file names
list.files(output,pattern = ".rds")

# Combine student files
combineStudents(inputdir = output, outputdir = output)

# Check file names
list.files(output,pattern = ".rds")

As we can see, a new file was created combining the achievement and the student questionnaire.

Combine schools

addSchools() adds school data to student and teacher files, using 'TIMSS' Advanced 1995:

# Check file names
list.files(output,pattern = ".rds")

# Add school data
addSchools(inputdir = output, outputdir = output)

# Check file names
list.files(output,pattern = ".rds")

unlink(file.path(tempdir(),"combineStudents"),recursive = TRUE)

As we can see, a new file was created combining the information of students and schools.

ILSAmerge documentation built on April 11, 2025, 5:54 p.m.