Gulliksen's pool: A quick tool for preliminary detection of problematic items in item factor analysis

Area of Science:

• Psychometrics
• Statistical analysis

Background:

• Exploratory factor analysis (EFA) is crucial for scale development, but problematic items can hinder accurate dimensionality assessment.
• Ineffective items complicate the early stages of scale construction, impacting the reliability of results.

Purpose of the Study:

• To propose and implement a procedure for flagging potentially problematic items before specifying a factorial solution in EFA.
• To enhance the item analysis process in scale development by identifying and excluding unsuitable items early on.

Main Methods:

• Developed a procedure based on item extremeness and consistency to define regions of item appropriateness and efficiency.
• Utilized extensive simulations to establish limits for appropriateness regions under realistic application conditions.
• Defined an Item Response Theory (IRT) index for prior item efficiency and proposed a combined approach for item subset selection.

Main Results:

• The proposed procedure effectively flags potentially problematic items, improving the accuracy of dimensionality and structure assessment in EFA.
• The method provides a robust way to select a more effective and problem-free subset of items for scale development.
• The procedure is particularly useful for normal-range measures and personality questionnaires.

Conclusions:

• The implemented procedure offers a valuable tool for researchers engaged in scale development, particularly when using EFA.
• This approach helps mitigate issues caused by inappropriate items, leading to more reliable and valid measurement instruments.
• The freeware implementation makes this advanced item analysis technique accessible to a wider research community.