Tests of Mean Differences Assignment
Due: Nov. 23, 2020
You are expected to complete this assignment on your own. Some of the questions require
multiple responses. Please make sure to answer all parts of a question.
For the Tests of Mean Differences Assignment Use the data file: EM777_moviemetadata2.sav or .xls
Remember to remove filters you might have on the data after going through Workshop 2.
1. Think about year the film was released [code: title_year]. (4 points)
a) Run a frequency report for year released. According to this data set, which year had the
most movie releases? How many movie releases did it have?
b) What is the range of years in this sample? Include the maximum and minimum from
the data to compute the range.
b) Perhaps you have the hypothesis: H1: Most of the movies were released after 1970.
Run a one-sample t-test to analyze this hypothesis. Please fill in the blanks:
t(__) = ____ , p = ____, d Cohen = ____.
c) Was this hypothesis (H A ) supported?
2. Consider visual representations of a t-test. You can insert pictures of your pen-and-paper
drawings, or you can use drawing software to create visuals. Do not insert pictures from the web
or book. (2 points)
a) Draw out a significant independent samples t-test using two normal distributions,
sample means, and confidence intervals. How do you know it’s significant (hint: is there
overlap in CIs)?
b) Draw out an independent samples t-test that is not significant using two normal
distributions, sample means, and confidence intervals. How do you know it’s not
significant (once again, hint: is there overlap in CIs)?
3. In your own words, discuss why ANOVAs are called “ANOVAs.” (1 point)
4. Let’s say we’re interested in whether the film was in black and white or color [code:
Color_0no] influenced IMDB’s score for the movie [code: imdb_score]. H2: There is a
significant difference between IMDB scores for movies in black and white and movies in color.
(6 points)
a) We should run an independent-samples t-test. Why? Also, explain why we can rule out
a paired-samples t-test, and a one-sample t-test?
b) Run the analysis. Should we assume equal variances? Why or why not?
c) We should use a two-tailed test. Why?
d) Use step 5.e on page 3 (H2) from Workshop 2 to write up the results. Please report the
means, standard deviations and confidence intervals around each group’s mean (not mean
difference). Don’t forget to use the codebook to understand the codes for Color_0no. (1
point for correct interpretation, 1 point for correct means, standard deviations and 95%
confidence intervals, 1 point for correct calculated effect size)
5. Run a paired-samples t-test comparing Actor 1’s Facebook likes [code:
actor_1_facebook_likes] to Actor 3’s Facebook likes [code: actor_3_facebook_likes]. (3 points)
a) Which variable had more Facebook likes?
b) Report the mean difference and confidence intervals around the mean difference.
c) Use the confidence intervals around the mean difference to discuss how we know
there’s a significant difference between actors without even looking at the p-value.
6. Run a one-way ANOVA using movie genre [code: Genre_Code] to predict IMDB score
[code: imdb_score]. Levene’s Test of Equality of Error Variances is significant. This means that
we should really run a different test. For this assignment, we’re going to ignore that. However,
note that this is an indication that one of our assumptions (homogeneity of variance) is broken.
(4 points)
a) Fill in the blanks:
F(__, __) = ____, p = ____, η2 = _____
b) Is movie genre a significant predictor of IMDB score? How do you know this?
c) Continuing to ignore that Levene’s is significant, use the Tukey’s HSD analysis to fill
in the first blank of each statement with either: “are” or “are not.” When the statement is
significant, fill in the second blank of each statement: > (greater than) or < (less than).
Tests of Mean Differences Assignment
(2 points – 1 for significance, 1 for directions)
IMDB scores for action movies ____ significantly ____ scores for adventure movies.
IMDB scores for action movies ____ significantly ____ scores for comedies.
IMDB scores for action movies ____ significantly ____ scores for dramas.
IMDB scores for adventure movies ____ significantly ____ scores for comedies.
IMDB scores for adventure movies ____ significantly ____ scores for dramas.
IMDB scores for comedies ____ significantly ____ scores for dramas.
Tests of Mean Differences Assignment