9.3 Lab Report Instructions

For this lab report, you will be analyzing real data and testing a hypothesis that is–as of yet–untested.

Import the data file into R:

mle <- readRDS("MLE.rds")

This data frame contains numerical ratings of ten emotions (anger, anticipation, disgust, fear, joy, sadness, surprise, trust, negative, and positive). It also contains a column with the category of each emoji.

Generate a hypothesis about how emoji differ in emotional content based on category. Your hypothesis should have one outcome variable that is one of the ten emotions. It should have one categorical predictor variable with three or more levels (that is, it should include three or more of the categories of emoji). Your hypothesis should include both an omnibus prediction and specific planned contrasts.

9.3.1 R Script

  1. Compute the M and SD of your outcome emotion for each level of your predictor variable.
  2. Conduct an ANOVA to test your hypothesis. You will need to convert the category variable to a factor. You may also need to subset your data.
  3. Conduct your planned contrasts.
  4. Produce a figure that vizualizes the differences between categories on your outcome variable.

9.3.2 Manuscript

Method

  • State your hypothesis clearly, including the groups and outcome variable you are using, your omnibus hypothesis, and your planned contrasts.
  • Briefly describe your analytic strategy.

Results

Report the following in your results section:

  • Number of emoji in each group included in your analysis.
  • M and SD on the outcome variable for each group.
  • Results of your ANOVA and planned contrasts.
  • Description of efforts to diagnose problems with statistical assumptions and data distributions.
  • A figure visualizing the relationship between your grouping and outcome variables.
  • The figure should display individual data points as well as a measure of central tendency (e.g., mean, median) for each group.

9.3.3 Discussion Questions

  • Provide a statement of support or non-support for your hypotheses (both your omnibus hypothesis and planned contrasts). Restate your findings in plain language (i.e., without reference to specific statistical tests or values).
  • Based on your analysis, are the Unicode-provided categories useful in explaining variation in emotional content of emoji? Your answer should reference both the effect size for your ANOVA and the spread of values displayed in your figure.
  • Describe one limitation of the study and how it may have influenced findings.
  • What is an implication of your analysis for research in this field, users of emoji, or industries/products related to emoji?