101689 Advanced Research Methods – Cheap Research Essays

Week 4 – Factor analysis
Part 1: Assumption checking and factor extraction

PREPARATION: • Read Chapter 22 of the Foolproof Guide to Statistics and review the relevant lecture
• Article provided on vUWS (Martin, Puhlik-Doris, Larsen, Gray, and Weir; 2003)
• Review the HSQ questionnaire on vUWS

Note that this exercise will not be included in the weekly exercises (Assessment 2). This tutorial IS designed to prepare you for the major project (Assessment 1). The questions listed below are a guide to the important aspects of the analysis, and what to expect of the class format. Analysis will be continued in week 5.

Scales are used extensively by practitioners and clinicians to diagnose psychological disorders and to measure the effectiveness of therapeutic interventions. They are also used by researchers to provide measures of psychological variables being studied. However, badly constructed scales can do more harm than good.

As such, an important skill you need to have as a psychologist is the ability to evaluate psychological scales before you use them. As an introduction to the scale evaluation process, you will be investigating the Humor Styles Questionnaire (HSQ).

Psychological research has identified sense of humour as an important buffer against the negative effects of stress. Basically, a good sense of humour can act as an effective coping mechanism. However, it is clear that people rely on different styles of humour, and not all styles of humour may have beneficial effects.

The HSQ was developed to identify individual differences in humour styles, with findings pointing to some distinct differences that may indeed have implications for health. Of course, for us to be confident with any results obtained from use of the scale, we must ascertain that it has strong psychometric properties.

ABOUT YOUR DATA SET

Participants

Data used in this study were collected as part of assessment in the third year psychology unit, Advanced Research Methods at the University of Western Sydney, between 2009 and 2015. The sample (N = 974, 581 female, 6 did not disclose; age 16-65, M = 26, SD = 9, 4 did not disclose) was drawn from students undertaking the unit and their acquaintances. Participation was voluntary and respondents remained anonymous. Detailed demographic information was not collected, but a diverse multicultural sample is assumed.

Materials

The Humor Styles Questionnaire (Martin, Puhlik-Doris, Larsen, Gray, and Weir; 2003), comprises four sub-scales, each with eight items, measuring two adaptive and two maladaptive dimensions of humour: (adaptive) affiliative, self-enhancing, (maladaptive) aggressive, self-defeating. Scores range from 1 (Totally Disagree) to 7 (Totally Agree), with a neutral mid-point, and 11 items (1, 7, 9, 15, 16, 17, 22, 23, 25, 29, 31) are reverse keyed. Scale reliability and validity have been established and found to hold strong within typical western multicultural settings.

THIS TUTORIAL
The main purpose of this week’s tutorial is to perform a principal components analysis (PCA) on the questionnaire data we have collected for Assessment 1 – The scale evaluation project.
Analysis will be continued in next week’s tutorial.

Step 1: Data screening and assumption checking

Due to limited time in the tutorial, reverse coding, assumption checking and data screening have been done for you. The following was done:

• Some questions were reverse coded – these are denoted with R next to them (for those interested please see Martin et al, 2003, for details).
• Univariate outliers were retained for analysis.
• From the initial sample of 974, 66 multivariate outliers were identified and deleted (resulting in a final sample of 908).
• While there were some issues with range and normality on a number of variables, overall, the test assumptions were met satisfactorily (but make sure you mention all relevant assumptions in your overview).
Step 2: Request analysis
Your Task Questions
Enter your answers to the questions below

(Note: If you find any of these questions challenging, READ the relevant chapter/s of the Foolproof Guide)

Analysis 1:
Run a Principal Components Analysis on all 32 HSQ variables, including.

(see p. 294 of the Foolproof Guide)

Ensure that you specify pairwise deletion of missing values.

In contrast to the foolproof guide, please note that you do not need to select the option for “Loading plots” in the “Rotation” tab.
For simplicity, items worded negatively have been reverse coded already.

This includes questions 1, 7, 9, 15, 16, 17, 22, 23, 25, 29, and 31. Question 1: What does the correlation matrix tell you?

Question 2: What is the KMO value, and is it acceptable?

Question 3: Do any variables appear to be outlying variables, in having very low communalities (< .30)? If so, which ones, and what are their communalities?

Question 4: How many components should be extracted using the eigenvalue > 1 criteria?

Question 5: How many components should be extracted using the scree plot criteria?
Question 6: How many components did we expect to find?)

Analysis 2:
Run a Principal Components Analysis forcing the number of components determined by theory.

For this analysis you should now remove a few options:

In “Descriptives”, remove “Initial solution”, “KMO”, and “Coefficients”.

In “Extraction”, deselect “Scree plot”, and change “Fixed number of factors” to “4”. Question 7: What are the eigenvalues for the factors extracted, and how much variance do they explain individually, and in total?

Question 8: Do any variables appear to be outlying variables in this analysis, in having very low communalities (< .30)? If so, which ones, and what are their communalities?
Week 3 – Reliability Analysis
PREPARATION: • Read Chapter 21 of the Foolproof Guide to Statistics
• Review the relevant lecture

Name:

Student number:

Each week consists of one tutorial exercise. These completed tutorials exercises (in addition to the SPSS output) are to be put into the relevant sections of this word document (or pasted at the end of the word document).

All students must submit their portfolios by 5pm Thursday 7th May. Online submission is through the Turnitin link (Assessment 2) in the Assessments tab. You will need to submit week 3 activity through the week 3 Turnitin link prior to 7th May. Please make sure that you submit this Week 3 Tutorial activity into the Week 3 Turnitin Link

Three weeks will be randomly chosen and marked. This contributes to your overall Assessment 2 mark.

One form of reliability is internal consistency reliability. This reflects whether items that are supposed to go together in a scale actually are answered in a similar way by respondents. In this exercise we’ll be looking at a measure of empathy, The Interpersonal Reactivity Index, to see how well items designed for its sub-scales fit together. Scale items are listed on the last page, and the four sub-scales include:
Fantasy (ability to have strong emotional experiences based on imagined events)
Perspective Taking (cognitive ability to see another’s point-of view)
Empathic Concern (emotional empathy for others)
Personal Distress (unpleasant feelings of distress in response to seeing another in distress)