Assignment 1: ANOVAs - Research Questions and Dataset

Assignment 1: ANOVAs

Participants were recruited to play Family Feud with research assistants.  They were told to compete for the high score by guessing the most common words related to a category word.  See below for example … they would not see the words until they got them correct:

Here’s a description of the dataset.

• Group: different types of Family Feud were given to participants
  • Without numbers: traditional Family Feud without the scores they would receive for getting it right
  • With numbers: Family Feud modified to show participants the scores they would receive if they got them right
  • Note that a given individual could only participate in 1 group or the other
• Helpers:
  • Singles: participants worked alone
  • Doubles: participants worked in pairs
  • Again, note that a given individual could only be involved in 1 of these 2 cases

• Total Score: participant’s final score in the game; higher scores indicate they did better (0-640).
• Columns D to L show scores (in %) for each participant broken down into different combinations of forward and backward word association as explained below.

• Forward: strength of relationship between words that you would normally be guessing in family feud (i.e. ashtray leading you to think of cigarette)
  • Levels: High, Medium, Low (referred to by the first part of variables H_, M_, L_)
• Backward: strength of reverse relationship (i.e. cigarette leading you to think of ashtray)
  • Levels: High, Medium, Low (referred to by the last part of variables _H, _M, _L)

The experimenters are interested primarily in 3 research questions:

• Do participants perform better or worse when they are prompted with the number they could win? Do participants perform better alone or as a pair? Are both of these factors independent of one another?
• Does performance depend on the forward strength of the words’ association, the backward strength, or a combination of both?
• Is the impact of forward and backward strength different for singles or doubles? (To clarify, we will ignore the “group” variable in this analysis. Also, focus on the novelty compared to Q2 in this analysis, i.e. don’t repeat anything already described through Q2).

Remember to paste your output from JASP or R in this document to complete this assignment.

1. Short description of the study variables (DV(s), IV(s), whether they are between-subject or within-subject variables, how many levels, how many conditions)
2. Data screening summary:
   1. Mention any missing data and how you checked for outliers  
   2. Assumptions (nature of the DV and normality)

Then, for each of these 3 research questions, do the following:

3. ANOVA
   1. Run the appropriate ANOVA
   2. Include the output
   3. Was there a problem of homogeneity of variance or sphericity?
   4. Write up the F statistics (for each main effect and interaction, with effect size)
   5. Post hoc tests with possible corrections (state the type of test, the type of correction, and include statistics written up, and output – no need for effect size here)
4. Plot the appropriate figure(s) with means and error bars showing 95% CI intervals, split by condition.
5. Write up a paragraph to pull together your findings and final conclusions on the research question. This requires group sizes, means, standard deviations or standard errors, statement of the type of ANOVA, the statistical results, all embedded in proper sentences like you would do in a research paper.