Examination for BSc Psychology (PRACTICE PAPER C)

4PAHPRM2 RESEARCH METHODS 2 PRACTICAL EXAMINATION

Section A

The questions in Section A are about the article by Suddendorf (2010).

You are provided with a replication of the experiment from this article in an SPSS data file named: Suddendorf Replication Specimen C.sav

Use the data in this file to answer the questions that follow.

Question A1 [8 marks available]

Suddendorf (2010) reports this analysis on page 494 of their article:

In response to the future question, 4-year-olds produced significantly more answers (M = 2.44, SD = 1.30) than 3-year-olds (M = 1.54, SD = 1.47), t = 2.94; p = .004.

Repeat this analysis for the data that you have been given, and report it.

(a) Table A1. Number of answers given to “Tell me something that you are going to do tomorrow” by age-group (report values to two decimal places)

(b) Report the t-test:

(i) The test statistic (to three decimal places): t = __________

(ii) The degrees of freedom df =  __________

(iii) The exact p-value (to three decimal places): p ____________

(Use ‘=’ or ‘<’ as appropriate to your answer.)

(e) Comparing your analysis in this question to the equivalent analysis shown above from Suddendorf (2010); the effect in my analysis is:

Select one answer.

Question A2 [10 marks available]

Suddendorf (2010) analysed the percentage of children who provided at least one answer to “Tell me something that you are going to do tomorrow” that was judged as ‘likely to be correct’. This is reported on page 494 of their article:

Less than half of the 3-year-olds generated an answer the parents judged to be likely correct (yesterday: 42%, tomorrow: 37%), whereas the majority of 4-year-olds did do so (yesterday: 63%, tomorrow: 81%). This age difference was significant for the future question (χ²(1, N = 82) = 16.27, p < .001) …

Repeat this analysis for the data that you have been given, and report it.

HINT: Use the variables named Age.group and Any.likely.tomorrow for this analysis

(b) The test statistic (to two decimal places): = __________

(c) The degrees of freedom df =  __________

(d) The sample size for this analysis N = __________

(e) Report the exact p-value (to three decimal places): p ____________

(Use ‘=’ or ‘<’ as appropriate to your answer.)

(f) Which kind of chi-square test did you use? (Select one answer):

Chi-square goodness-of-fit test / Chi-square test for contingency tables

(g) What do you conclude from this test? (One sentence is sufficient)

Question A3  [7 marks available]

Suddendorf (2010) reports this analysis on page 494 of their article:

As predicted, there was a significant positive correlation between number of likely correct answers on the yesterday and the tomorrow questions (r = .49, p = .001) and this association remained significant when age was partialled out (r = .46, p = .001).

Repeat this analysis for the data that you have been given, and report it.

Report the correlation on this page, and the partial correlation on the page that follows

(a) Obtain and report the correlation (r) between the number of likely correct answers on the yesterday question and the number of likely correct answers on the tomorrow question:

(i) The correlation coefficient (to three decimal places): r = __________

(ii) The degrees of freedom for r: df = __________

(iii) The exact p-value (to three decimal places): p _____________

(Use ‘=’ or ‘<’ as appropriate to your answer.)

(b) Obtain and report the partial correlation that Suddendorf reports above, by partialling out the variable exact.age from the correlation that you obtained in part (a).

(i) Partial correlation (to three decimal places): r_p = __________

(ii) Degrees of freedom for r_p: df = __________

(iii) Report the exact p-value (to three decimal places): p ____________

(Use ‘=’ or ‘<’ as appropriate to your answer.)

(c) From your analysis reported in parts (a) and (b) above:

Select one answer.

Question A4 [9 marks available]

(a) Which of the following research terms accurately describe some aspect of the design of Suddendorf (2010)?

Select ‘Yes’ or ‘No’ for each term.

(b) Justify the answers that you gave in part (a) above. The equivalent of one to two well-written sentence per part (i to iii) will be sufficient for this question.

Section B

The questions in Section B are about the Cups Game Study

Use this data file for Section B: Specimen C Game.sav

Question B1 [7 marks available]

This question is about methodological hypotheses M2a and M2b:

M2a The time taken (in seconds) to read and answer the pre-game question will have a similar distribution in each condition.

M2b The average time taken (in seconds) to read and answer the pre-game question will not differ significantly between the two conditions.

(a) The 95% confidence interval (CI) for the mean difference between the two conditions for the time taken to read and answer the pre-game question conditions can be used to examine hypothesis M2b.

(i) Use SPSS to obtain that CI, and report it (to two decimal places).

95% CI for the mean difference in the time taken to read and answer the pre-game question between the two conditions:

Lower limit: __________ seconds Upper limit: __________ seconds

(ii) Delete/select as appropriate:

The 95% CI for the mean difference reported in (c)(i) above is / is not consistent with hypothesis M2b.

(iii) Which feature of the confidence interval leads you to that conclusion?

(b) The researcher has put forward hypotheses M2a and M2b because the interpretation of any differences between conditions (e.g., predictions P3, P4 or P5) depends upon whether or not the data are consistent with M2a and M2b. Briefly explain how the interpretation of the effects predicted in P3, P4 or P5 is affected by the considerations described in M2a and M2b.

Two or three well-written sentences are sufficient for this question.

Question B2 [17 marks available]

This question is about research question R8:

R8 For each condition considered separately, does the time taken to read and answer the pre-game question (Time_Pre.Qn) predict the time taken to complete the Cups Game (Time.Game)? In this respect, how do the two conditions compare with each other?

Hint: For this question, you will need to use a filter, or split the file, to examine each condition separately. Remember to turn the filter off, or unsplit the file, when you have finished this question.

(a) Use simple linear regression to examine research question R8.

Undertake the analysis in SPSS that will allow you to write down the regression equation, for each condition (separately), that can be used to predict the time taken to complete the Cups Game from the time taken to read and answer the pre-game question.

When you do so, obtain the 95% confidence interval for the regression coefficients. Do NOT use the ‘Bootstrap’ function for this.

Record the results of your analyses in the table below

Table B2. Linear regression predicting the time taken (seconds) to complete the Cups Game from the time taken (seconds) to read and answer the pre-game question, shown for each condition (report values to three decimal places)

(b) Complete/Delete as appropriate:

(i) For the deliberation condition, the direction of the linear relationship is _____________ and this relationship IS / IS NOT statistically significant.

(ii) For the control condition, the direction of the linear relationship is _____________ and this relationship IS / IS NOT statistically significant.

Question B2 (continued)

(c) For the deliberation condition, write down the equation for the regression line. Use G to represent the time taken (in seconds) to complete the Cups Game, and use P to represent the time taken (in seconds) to complete

Equation: ________________________________________________

(d) Imagine that someone was shown your completed version of Table B2 (on the previous page) but WITHOUT the information in the final column on the right. Briefly explain how this individual could use the information in some of the other columns of Table B2 to determine whether each of the linear relationships reported in the table was statistically significant.

Two or three well-written sentences are sufficient for this question.

Reminder: If you have used a filter for question B2, turn the filter off before starting your next question. If you have split the file to answer question B2, unsplit the file before starting your next question.

Question B3 [27 marks available]

For this question, you need only consider the following two predictions:

P3 The average number of times that the red cup is picked will be significantly greater in the deliberation condition than in the control condition.

P5 The proportion of participants using the maximising strategy will be significantly higher in the deliberation condition than in the control condition.

Use the data that you have been provided with to write a Results section that would be appropriate for a research report in psychology. Analyse and report only the information that is appropriate and relevant to the two predictions listed above. You should assume that the only variables of interest for your Results section are the ones needed to investigate these two predictions.

You may use tables and graphs where necessary or beneficial. Marks will be awarded for layout, clarity and correct interpretation of the statistical data, and appropriate description of how the data were treated. In order to save space and time, you may use the abbreviations P3 and P5 to refer to the predictions. Any other abbreviations that you use should be defined first. You may assume that parametric statistical tests can be used for analyses that involve measurement data (i.e., for analyses that involve at least one NON-categorical variable). Report a t-test if you are analyzing a mean difference. Exact p-values are preferred to reporting ‘levels’ via the “>” or “<” symbols, except where p < .001.

Write your Results section in the space provided below.