PART I (22 points)
A study by Chen and colleagues (2014) tested what is known as the ‘beer-goggles effect.’ This is the idea that subjective perceptions of physical attractiveness becomes less accurate after drinking alcohol. The logic is that: (i) alcohol consumption reduces accuracy in symmetry judgements, and (ii) symmetrical faces (like Denzel Washington) are rated as more attractive. Thus, drinking alcohol will lead people to rate unattractive people as being more attractive than they actually are (based on facial symmetry) – the ‘beer-goggles effect.’ However, drinking alcohol should not impact attractiveness ratings for people who are initially highly attractive (based on facial symmetry). The data below is not the actual data from the Chen et al. (2014) study.
There are 32 participants who were randomly assigned to one of three Alcohol Conditions:
(i) A Low-dose group who drank 500ml of low-strength beer (n = 16)
(ii) A High-dose group who drank 500ml of high-strength beer (n = 16)
After their drink, participants rated the attractiveness of 50 faces that had been pre-assigned (based on symmetry) to one of two Face Types: Attractive or Unattractive. Thus, half of the participants within a group (n = 8) received 50 attractive faces and the other half (n = 8) received 50 unattractive faces.
The attractiveness of the faces was rated on a 0 to 10 point scale (higher numbers = more attractive) and averaged together for each participant. This Attractiveness Rating is the dependent variable.
Round all calculations to 2 decimal points as you conduct them. This will help to ensure that round-off error is the same for everyone (including for the grading TAs).
Use this scenario to answer the questions in Part 1.
Face Type Low-dose Group
n = 16 High-dose Group
n = 16
n = 16 M = 6.63
SD = 0.74 M = 6.75
SD = 1.28
n = 16 M = 4.25
SD = 1.39 M = 6.50
SD = 1.69
- Define what a main effect is, then state what the main effects are in this example. (2 points)
- Define what an interaction effect is, then state the interaction in this example. (2 points)
- Use the Ms in the table above to answer the following questions. (3 points total)
a. Calculate the two means that should be compared to consider whether there is a possible main effect for Alcohol Condition? [Hint. The means are not in the table. Show your calculations.] (1 points)
b. Calculate the two means that should be compared to consider whether there is a possible main effect for Face Type? [Hint. The means are not in the table. Show your calculations.] (1 points)
c. Which means should be compared to consider whether there is a possible interaction between Alcohol Condition and Face Type? (1 points)
- The calculation questions below use the example study and data above. The question is broken down into sub-questions. Please make sure to complete the calculations within each sub-question so that you can receive partial credit for your calculations that are correct, if needed. Show your work. [Hint. A few values are already filled in in the ANOVA Table below to help with the number of calculations that will be needed. It may be helpful to follow along with the lecture to make sure you’re following the steps correctly.] 13 points total)
a. Compute your MSwithin and MSbetween. (2 points)
b. Convert MSbetween to SSbetween. (1 points)
c. Compute the MSalcohol and SSalcohol. (2 points)
d. Compute the MSface and SSface. (2 points)
e. Compute the SSIXN. (1 point)
f. Compute the MSIXN. (1 point)
g. First, restate your MSwithin, MSalcohol, MSface, and MSIXN from the previous questions. Then compute your F-ratios. You MUST show your work for the F-ratio calculations. (3 points)
h. Correctly fill in the ANOVA Table. The values given to you below are so you can check your work, so if your values are slightly different due to rounding error, change the values below to what you calculated. If your values are very different, recheck your calculations. (1 point)
Source SS df MS F-ratio
Face Type Condition
Alcohol x Face Type 5.15
Total 82.80 31
- Calculate the effect size (eta-squared) for the interaction effect, and interpret what it means in the context of the question. (2 points)
For Part II of the Assignment this week you will actually be reporting on the analyses that you conducted in lab section, rather than conducting a new set of analyses. Thus, you will not be copy and pasting R code below. Instead, you will focus on interpreting the output.
- Interpret the results from the two-way ANOVA that you conducted in Lab 4 this week below. Ensure that you are including all of the needed components in your interpretation. Your results must be written up in APA Style. For full credit, you must report results for all effects, whether they are statistically significant or not. Include descriptive statistics for all effects as well. (10 points)
- In lab section, you created a line graph with Condition on the x-axis, and Major represented as different lines. Create a line graph instead with Major on the x-axis and Condition on separate lines. (2 points)
a. Paste the R code below. (1 point)
b. Paste the line graph below. (1 point)