PSY137 – Personality 2 Confirmatory Factor Analysis • It is used to test whether measures of a construct are consistent with a researcher's understanding of the nature of that construct (or factor).
• the objective of confirmatory factor analysis is
to test whether the data fit a hypothesized measurement model. Confirmatory Factor Analysis • the researcher first develops a hypothesis about what factors s/he believes are underlying the measures s/he has used.
• If you intend to come up with a scale on
personality dispositions, and you hypothesize that there are 3 unrelated concepts under this, then you can create a model to show this. Steps 1. Choose a model / theory
• Choose a theory or a hypothesis within that
theory as the basis for coming up with your scale. Steps 2. Come up with statements that would correspond to each domain of that construct.
• E.g. Moving Toward People – I feel the need to
be approved of constantly. • Moving Against People – I feel the need to criticize every person I meet. • Moving Away from people – I don’t like being too close to people. Steps 3. Come up with 100 statements all in all.
• If you have 4 domains under that construct, then
you should come up with 25 items per domain. • It would be helpful to have equal items for all domains. • This would ensure that the test encompasses the construct being measured. • Make sure the test will have a likert scale format. Steps 4. Have the test with the items validated by an expert.
• Prepare a short description of the theory you
are using. • Define each domain you intend to measure in your personality test. • The validator will suggest revisions on how to develop your scale better. Steps 5. After revising the test according to the validator’s specifications, administer it to no less than 200 participants.
• Make sure the participants are representative
of your intended population. • Having 200 participants will have a better chance of arriving at significance. Steps 6. After pilot testing, plot the scores on excel. Steps 7. Create a covariance matrix for your data on excel. - get the mean scores per domain. - click data analysis - click covariance - type ‘Means’, ‘Std. Dev.’, ‘Cases” (how many participants), and ‘Matrix’ (4) Highlight only the values, not the case numbers (first column), then click ok. SOCIAL PHYSICAL EMOTIONAL ACADEMIC SOCIAL 0.154221203 PHYSICAL 0.086914231 0.15630654 EMOTIONAL0.089233513 0.07328899 0.13642979 ACADEMIC 0.099344836 0.08183363 0.08541742 0.2304841 Means Std. Dev. Cases 445 Matrix 4 Steps 8. Subject your covariance matrix to statistica. Open the file (covariance matrix) Click on case names for column and rows, then click ok. Click “statistics”, then Advanced Linear/NonLinear models, then structural equation modeling Click “path wizards” Click “confirmatory factor analysis” then click “ok”. Type your latent variables then click on “vars” then choose 1 domain in “select manifest variables” until all domains are placed. After all domains are included in manifest variables, click OK. Click “append this model to existing program” then click ok. It should look like this. Click “ok (run model) It should appear this way. Click “ok” It should look like this. Then click “summary”. It should look like this.Click “structural equation below (indicated). Click “advanced” Click on “Goodness-Of-Fit Indices” then click “non-centrality-based indices” Evaluating Model Fit • several statistical tests are used to determine how well the model fits to the data.
• Note that a good fit between the model and
the data does not mean that the model is “correct”, or even that it explains a large proportion of the covariance.
• A “good model fit” only indicates that the
model is plausible. Cronbach Alpha • coefficient of internal consistency. – It measures whether several items that propose to measure the same general construct produce similar scores.
– For example, if a respondent expressed agreement
with the statements "I like to ride bicycles" and "I've enjoyed riding bicycles in the past", and disagreement with the statement "I hate bicycles", this would be indicative of good internal consistency of the test.
• It is commonly used as an estimate of
the reliability of a psychometric test for a sample of examinees. Steps 9. Compute for the Cronbach Alpha. Click statistics, then “multivariate exploratory analysis”, then “reliability/item analysis”/ Click “variables”, then click “select all”, then click “ok” Click “ok” Clicl “summary: item-total statistics” Check the cronbach alpha. If it is .80 and higher, it means the scale you developed is reliable and consistent. Requirements 1. Background of the Theory 2. Definition of Construct and sub-variables you intend to measure. 3. 1st draft of the test (prior to validation) 4. Validated test instrument 5. Excel sheet of score 6. Statistica printscreen of CFA 7. Statistica printscreen of Cronbach Alpha