fsia-package: Import and Analysis of OMR Data from FormScanner

In fsia: Import and Analysis of OMR Data from FormScanner

Description

Import data of tests and questionnaires from FormScanner. FormScanner is an open source software that converts scanned images to data using optical mark recognition (OMR) and it can be downloaded from <http://sourceforge.net/projects/formscanner/>. The spreadsheet file created by FormScanner is imported in a convenient format to perform the analyses provided by the package. These analyses include the conversion of multiple responses to binary (correct/incorrect) data, the computation of the number of corrected responses for each subject or item, scoring using weights, the computation and the graphical representation of the frequencies of the responses to each item and the report of the responses of a few subjects.

Details

Package:	fsia
Type:	Package
Version:	1.1.1
Date:	2017-06-23
License:	GPL-3

Data of questionnaires and tests are often collected on paper forms. FormScanner is an open source software that converts scanned images to data using optical mark recognition (OMR). Function read.formscanner of the fsia package can be used to import data from FormScanner in R. The correct response (key) can be specified using function addkey. It is also possible to specify weights for each response using function addweights. If items have a key, data can be converted to binary variables using function resp2binary. In this case, the number of corrected responses for each person and for each item can also be computed by using functions person.stat and item.stat. These functions can also be used to compute a score using the weights previously specified. Function freq calculates the absolute or percentage frequencies of the responses to each item. The frequencies can be printed on screen or plotted on a graph. In both cases, the true responses (if any) are highlighted. The responses given by one or a few subjects can be displayed on a graph by using function report. The key is shown on the right and wrong responses can be immediately identified by the red colour.

The package includes two data sets for illustrative purposes. Data sets test and questionnaire contain the result of importing csv files with function read.formscanner. Data set key contains the correct responses of the items of the test data sets. Data set weights contains the weights of each correct response, while data set weights_multiple contains the weights of each response.

Author(s)

Michela Battauz

Maintainer: Michela Battauz <[e-mail]>

References

Borsetta, A. (2017). FormScanner [Computer Software], URL http://sourceforge.net/projects/formscanner/.

Examples

# IMPORT DATA FROM FORMSCANNER
# find the directory with package fsia
dir_pkg <- find.package("fsia")
# the example files are in the directory examples
# create the path
questionnaire_path <- file.path(dir_pkg, "examples", "scan_results_questionnaire.csv")
test_path <- file.path(dir_pkg, "examples", "scan_results_test.csv")
# import file "scan_results_questionnaire.csv"
questionnaire_imp<-read.formscanner(questionnaire_path, dummy = "Q5.sources")
questionnaire_imp
# questionnaire_imp is equal to the data questionnaire
# import file "scan_results_test.csv"
test_imp <- read.formscanner(test_path, conc = paste("id", 1:6, sep = ""), id = "id1")
test_imp
# test_imp is equal to the data test

# ADD THE KEY
# create the path for file "key.csv"
key_path <- file.path(dir_pkg, "examples", "key.csv")
# add the key 
testk <- addkey(test_imp, keyfile = key_path)
testk$key

# ADD WEIGHTS
# create the path for file "weights.csv"
weights_path <- file.path(dir_pkg, "examples", "weights.csv")
# specify the weights for each correct response
testw <- addweights(testk, weightsfile = weights_path)
testw$weights
# create the path for file "weights_multiple.csv"
weights_mult_path <- file.path(dir_pkg, "examples", "weights_multiple.csv")
# specify the weights for each response
testwm <- addweights(test_imp, weightsfile = weights_mult_path)
testwm$weights

# CONVERT DATA TO BINARY VARIABLES
resp01 <- resp2binary(obj = testk, col = 2:41)
resp01[, 2:5]

# ASSIGN WEIGHTS TO RESPONSES
resps <- resp2scores(obj = testw, col =2:41)
resps[, 2:5]

# ASSIGN WEIGHTS TO RESPONSES (MULTIPLE WEIGHTS)
resps <- resp2scores(obj = testwm, col =2:41)
resps[, 2:5]

# PERSON STATISTICS (selected only 4 items)
pst <- person.stat(obj = testk, col = 2:5)
pst
pst <- person.stat(obj = testw, col = 2:5, weights = TRUE)
pst
pst <- person.stat(obj = testwm, col = 2:5, weights = TRUE)
pst

# ITEM STATISTICS
ist <- item.stat(obj = testk, col = 2:41)
head(ist)
ist <- item.stat(obj = testw, col = 2:41, weights = TRUE)
head(ist)
ist <- item.stat(obj = testwm, col = 2:41, weights = TRUE)
head(ist)

# FREQUENCIES OF THE RESPONSES
fr <- freq(obj = testk, col = c("Question03", "Question04"))
fr
par(mfrow = c(1, 2))
plot(fr, ask = FALSE)

# RESPONSES OF TWO SUBJECTS
par(mfrow = c(1, 2))
report(obj = testk, col = 2:11, whichid = c("102344", "245784"))
report(obj = testw, col = 2:11, whichid = c("102344", "245784"), weights = TRUE)
par(mfrow = c(1, 1))
report(obj = testwm, col = 2:11, whichid = c("102344", "245784"), weights = TRUE)

fsia documentation built on May 2, 2019, 5:42 a.m.