A psychologist is interested in Instagram ‘influencers’ and how many likes their most popular video receives. They hypothesise that four variables might contribute to the number of likes: the length of the video in seconds; the perceived attractiveness of the influencer; the sex of the influencer; and whether a dog is present in the video.

The psychologist conducts a multiple regression to predict the number of likes for the most popular video.

There are four predictor variables Length_of_video (in seconds) Dog_present (0 = no, 1 = yes) Attractiveness (0 – 10, higher number = greater rated attractiveness) Sex (0 = male, 1 = female)

SPSS output from the study is provided. Use the output to answer the questions:

Page 2 of 14 Page 3 of 14 Question 1a:

Examine the frequency tables for ‘Sex’ and ‘Dog_present’. The psychologist has made an error in their data entry. What error have they made, and how might this impact the analyses if it was not fixed? [3 marks]

Question 1b:

The psychologist spots and fixes the error. How much variance is explained in video likes by the four predictor variables? [2 marks]

Question 1c:

Write up the complete regression model in APA format, providing standardised (beta) coefficients. [10 marks]

Question 1d:

Calculate and provide the cut off value for Cook’s distance. [2 marks].

Question 1e:

Should the psychologist be concerned by influential cases in their analyses? Explain your answer. [3 marks]

Page 4 of 14 Question 2: Between Subjects ANOVA

A psychologist is interested in road-rage experienced by drivers of different types of vehicles (Bicycle, Motorcycle, Car, Van, Bus). They develop a road-rage scale which measures the average number of times a curse-word is uttered per 60 minutes of journey taken. More curse-words are indicative of greater road-rage.

The psychologist was conducting exploratory research and had no specific hypotheses regarding which drivers experienced the most road-rage.

SPSS output from the study is provided. Use the output to answer the questions:

Page 5 of 14 Page 6 of 14 Page 7 of 14 Question 2a:

Write up the full analysis and appropriate follow up tests. [12 marks]

Question 2b:

How much variance did the groups predict in road-rage score? [2 marks]

Question 2c:

If the psychologist decided not to conduct an ANOVA, but to conduct a series of t-tests, how many would they need to perform to compare all groups? [2 marks]

Question 2d:

In the Contrast Coefficient table above, what comparison would contrast 1 be testing? [2 marks]

Question 2e:

The psychologist has made an error in defining their contrasts. Explain what the error is. [2 marks]

Page 8 of 14 Question 3: Complex between subjects ANOVA

The current results are from a study exploring the influence of social media use on self-esteem, in male and female students.

Participants were recruited if they used one of the following forms of social media (and none of the others):

1- Facebook

2- Twitter

3- Instagram

This gave a three level between subjects factor of ‘social media’. Next, we recorded participants’ sex (a two level between subjects factor, coded as male or female). Finally, participants completed a self-esteem scale, in which higher scores are indicative of better self-esteem.

The design of the study was therefore a 3 x 2 between subjects design, with “Social media” as the first between subjects’ factor (Facebook / Twitter / Instagram) and “Sex” (Male / Female) as the second between subjects factor. The DV is self-esteem.

It was hypothesised that Instagram use will reduce self-esteem, and this reduction will be greater in females.

SPSS output from the study is provided. Use the output to answer the questions:

Page 9 of 14 Page 10 of 14 Page 11 of 14 Page 12 of 14 Page 13 of 14 Question 3a:

Produce an APA formatted descriptive statistics table. [5 marks]

Question 3b:

Do the data have homogeneity of variance? Explain your answer. [2 marks]

Question 3c:

Write up the main effects from the analysis, reporting post hoc tests where necessary. [6 marks]

Question 3d:

Write up the interaction, reporting post hoc tests where necessary. [5 marks]

Question 3e:

Do the results support the hypothesis? Briefly explain your answer. [2 marks]