## t(64) = -2.11, p = .039. Calculation of Cohen’s d indicated the presence of a small effect size, d =.26, 95%CI [.01, .50]. ? Draw a conclusion. Is the effect potentially important? QUESTION 5 (Needs bibliography review) Explain the advantage of reporting effect sizes in addition to significance levels. (20 marks) ? What is statistical significance? ? What is effect size?

INSTRUCTIONS ON ASSESSMENT:
QUESTION 1
A researcher is interested in the relationships between the following four variables: On-line game
addiction, narcissistic personality traits, self-control, and aggression. Data are collected from a
sample of 226 individuals. The data can be found in the assessment SPSS spreadsheet
‘assessment.sav’. For each variable, higher values indicate greater levels of the variable. Analyse
the data using SPSS and report the results using standard APA conventions. (20 marks)
? The focus is clearly on the relationships between
– Narcissistic personality traits
– Self-control
– Aggression
? State what you are going to do to investigate this
– Descriptives
– Scatterplots
? Pearson Correlations Produce a table of descriptive statistics
– Mean
– Standard Deviation
– Skew
– Kurtosis
– N
? Comment on any observed deviations in skew or kurtosis (Z-scores that are greater
than +/- 1.96) but indicate that no transformations will be performed. Indeed, there are
few options for kurtosis and it is not generally seen as a big problem anyway.
? Present scatter plots between all pairs of variables with lines of best fit added.
– Could be individually or in a matrix.
– Comment on what the relationships look like.
– Do there seem to be any outliers, and if so, do they look like they are going to make an unacceptable impact on the relationship observed?
? Present correlations matrix.
– Comment on the strength, significance and direction of the correlations.
QUESTION 2
Following the initial analysis conducted in Q1 the researcher goes on to look at which of the
variables (On-line game addiction, narcissistic personality traits, and self-control) are the best
predictors of aggression. Analyse the data using SPSS and report the results using standard APA
conventions. (30 marks)
? All needs to be referenced according to APA style.
? It requires multiple regression
? Briefly indicate whether the variables would be expected to be good predictors based on correlations reported in question 1
? R Squared
? Anova – significance and what it means
? Individual Predictors
o Beta coefficients and confidence intervals
o T-tests for the beta coefficients § Part correlations (compare to zero order)
o Collinearity
? Indicate which appears to be best predictor, second, third.
? State the regression equation
QUESTION 3
The researcher also thinks that it is possible that there are sex differences for these variables. You
will also find codes for males and females in the data set. Analyse the data using SPSS to investigate
this, and report the results using standard APA conventions. (20 marks)
? All needs to be referenced according to APA style.
? State that you performed Independent Sample t-tests to identify if sex differences exist.
? Table of descriptive statistics
– Mean, SD, N
? Report each t-test separately:
EXAMPLE:
The 31 participants in the rosemary condition (M = 15.90, SD = 2.45) performed significantly better on time based prospective memory than the 34 participants in the no aroma condition (M = 14.41, SD = 2.451), t(63) = – 2.59, p = .012. Calculation of Cohen’s d indicated the presence of a medium effect size, d = .64, 95% CI [.14, 1.14].
? Draw an overall conclusion. Are there sex differences? Which variables? How big? Consistently in the same direction?
QUESTION 4
As I further investigation, the researcher wants to see if playing a violent video game increases
levels of aggression. She runs a study in which aggression levels are measured in 40 participants
before and after playing a violent video game. Use SPSS to compare the variables Aggro_pre and
Aggro_post in order to answer this question. (10 marks)
? All needs to be referenced according to APA style.
? State that you performed a Paired samples t-test to see if aggression after playing the game is greater than before.
? Table of descriptive statistics
o Means, SD, N
o Report t-test.
EXAMPLE:
The results from the time based prospective memory test (M = 15.12, SD = 2.42) and the event based prospective memory test (M = 15.75, SD = 1.64) revealed significantly better performance for event based prospective memory, t(64) = -2.11, p = .039. Calculation of Cohen’s d indicated the presence of a small effect size, d =.26, 95%CI [.01, .50].
? Draw a conclusion. Is the effect potentially important?
QUESTION 5 (Needs bibliography review)
Explain the advantage of reporting effect sizes in addition to significance levels. (20 marks)
? What is statistical significance?
? What is effect size?
? The distinction between statistical significance and scientific (or practical)
importance.
? Give an example.
? Factors that affect significance
? Problem with effect sizes eg many different types
? Value of effect size in addition to statistical significance
? Overall summary/conclusion
? All needs to be referenced according to APA style.