Psychometric analyses of a personality test: Is the test psychometrically acceptable? Does neuroticism

Measurement and Assessment (M&A) II Assignment (resit)

This M&A II assignment consists of carrying out psychometric analyses in SPSS, and reporting the results. This is an individual assignment, so you are required to do the analyses and write the research report by yourself. The research report should be comprehensible for people who are not familiar with the details of psychometric analyses.

Below the required analyses are detailed. You can do all these analyses in SPSS. In reporting the results, you can cut-and-paste SPSS tables, if you like, but tables should include your own table caption (titles) and table number and you should remove non-relevant information from the tables. The report should include clear statements on the reason for the analyses (what are the questions?) and the meaning of the results (what are the answers to the questions?).

Word limit

The word limit 1500 words (not including tables)

Writing and plagiarism checking

The report should be written in clear and correct English. Make sure your report contains no spelling or grammatical errors (please run a spelling check). You are required to do this assignment individually and to write the report individually. Note: all reports will be submitted to a plagiarism checker.

Assignment: Extroversion, neuroticism & employment

In a pilot study of the psychological consequences of unemployment among young adults, a short personality test was administered to a sample of N=407 participants between 18 and 22 years. The test measures the personality traits neuroticism and extroversion by means of 5 neuroticism and 5 extroversion items. In addition, the participants were asked whether they were gainfully employed. The answer to this question is coded 0 for employed and 1 for unemployed. This variable is called work. The aim of the pilot study was to determine whether:

1) the personality test is psychometrically acceptable

2) neuroticism and extroversion predict the variable work

The data. The sample comprises 407 young males and females between 18 and 22 years. The 5 neuroticism items are denoted fear, annoyance, depression, shame, vulnerable; the 5 extroversion items are denoted cordial, social, energy, adventurous, joyous. Each item contains a statement such as “Generally, I am a joyful person”. The response format is a 7 point scale, with score 1 indicating “disagree completely” and 7 indicating “agree completely”. A higher test score indicates a greater degree of neuroticism or extroversion. In addition, as mentioned above, there is the dichotomous variable work (0 means “employed” and 1 means “unemployed”). The SPSS system file containing the data is called NeurotExtrav.sav.

The analyses (SPSS menu navigation is discussed below). The present aim is twofold.

The first aim is to investigate the following psychometric characteristics of the personality test:

Dimensionality & Item quality. Carry out a factor analysis of the 10 items. Fit the two common factor model to the data, and report, following oblique rotation, the factor loadings, the communalities, the factor correlation, the percentage of variance explained by each of the two factors, and the total percentage of variance explained in the two factor model. Firstly, given that the test is supposed to measure the two latent variables neuroticism and extroversion, are the results interpretable as such? Secondly, are the test items good indicators for neuroticism and extroversion? Recall from Workbook CH3 that: “in general the items with a loading <.30 are considered as items that do not fit well within the set of questions, the questionnaire, or the factor.”

The second aim is to analyze the relationship between personality and unemployment.

Logistic regression analysis. Determine whether neuroticism and extroversion predict the employment status (work). To this end, first calculate the test scores by summing the neuroticism into a single score and by summing the extroversion items into a single score. Calculate the z-scores of the test scores by standardizing the test scores. Use the z-scores in the logistic regression analyses as the predictors. Do the analysis twice: once with the neuroticism z-scores as the predictor, and once with the extroversion z-scores as the predictor. Include in your report the relevant null hypotheses, and alpha level of 0.05. Do neuroticism and extroversion predict the probability of being unemployed (work=1)? If so, report 1) the unconditional probability of work=1, and 2) the probability of work=1 given the mean of the z-scores (zero) and 3) given the mean plus one standard deviation of the z-scores. Because the predictor is standardized, the mean equals zero and the mean plus 1 standard deviation equals 1. So in your discussion of the effect size, include

prob(work=1), the unconditional probability, and the following conditional probabilities

prob(work=1|zscore=0) = 1 / (1 + exp(-(b0 + b1*0)) and

prob(work=1|zscore=1) = 1 / (1 + exp(-(b0 + b1*1)).

Discuss, based on your subjective evaluation of these probabilities, the strength of predictive value of the personality traits.

Report sections

Carry out the required analyses in SPSS (instruction are given below), and write a report on the basis of the results. Your report should include:

1) Title page. Title page with a sensible title, your name and student number, and the word count (make sure that the word count is < 1500).

2) Introduction. An introduction explaining the aims of the study.

3) Method section. A method section in which the methods are discussed briefly – which analyses are carried out (factor analysis and logistic regression analyses) and which questions do these answer? In the case of the logistic regression analyses, mention the null hypotheses and the alpha level of 0.05 (i.e., a=.05).

4) Data section. Brief description of the variables and the samples (including information on sex and age).

5) The results (4 tables). Provide descriptives (table 1 with means and standard deviations of the items and the variables work, sex and age, table 2 with the correlation matrix of the items).

5A) Describe results of the factor analysis (table 3 with rotated factor loadings and communalities). Report the correlations between the common factors and the percentages of explained variance after rotation (for each factor and for both factors together) in the text.

5B) Describe the results of the logistic regression analyses (table 4 with the parameter estimates, standard errors and p-values).

6) Conclusions. Conclusions with a clear answer to the question concerning the psychometric aspects of the test and the predictive value of the neuroticism and extraversion test scores. Is the test acceptable in terms of interpretation of the factor structure and is the quality of test items adequate? Do the personality traits predict the variable work? If so, what is the effect size, as expressed in terms of conditional probabilities? (see above for explanation).