Scribd
Upload
34 views

Exercise 5 Answer Key

Model 1 and Model 2 estimate the effect of movie ratings and genres on consumers' likelihood to watch movies. Model 1 finds a uniform effect of ratings across genres, while Model 2 allows the rating effect to vary between genres through interaction terms. Based on Model 1, a pure adventure movie would need a IMDB rating 1.24 points higher than a pure romance movie to have the same likelihood of being watched.

Uploaded by

argirobri
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
34 views

Exercise 5 Answer Key

Model 1 and Model 2 estimate the effect of movie ratings and genres on consumers' likelihood to watch movies. Model 1 finds a uniform effect of ratings across genres, while Model 2 allows the rating effect to vary between genres through interaction terms. Based on Model 1, a pure adventure movie would need a IMDB rating 1.24 points higher than a pure romance movie to have the same likelihood of being watched.

Uploaded by

argirobri
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Exercise 5 Answer Key

Coefficientsa

Model Unstandardized Coefficients Standardized t Sig.


Coefficients

B Std. Error Beta

(Constant) 1.176 .039 30.239 .000

comedy .868 .024 .427 35.539 .000

adventure .215 .024 .106 8.810 .000


1
romance .333 .024 .164 13.635 .000

rating .095 .005 .245 20.431 .000

days -.005 .001 -.052 -4.340 .000

a. Dependent Variable: likelihood

Coefficientsa

Model Unstandardized Coefficients Standardized t Sig.


Coefficients

B Std. Error Beta

(Constant) 1.366 .055 24.812 .000

comedy .581 .049 .285 11.793 .000

adventure .239 .049 .117 4.853 .000

romance .228 .049 .112 4.635 .000

1 rating .055 .009 .142 5.807 .000

days -.005 .001 -.052 -4.318 .000

comedy_rating .062 .009 .180 6.716 .000

adventure_rating -.006 .009 -.017 -.647 .518

romance_rating .023 .009 .066 2.434 .015

a. Dependent Variable: likelihood


1. Answer this question based on both the models (Model 1 and Model 2). If the findings are
different across the models, explain why they are different. If a question cannot be answered
using a model, provide the reason for it.

What is the effect of 1 unit increase in IMDB rating on the consumers’ likelihood to watch the
following types of movies?

Please notice that the question asks for “a 1-unit increase in rating”: it does not mention
anything about the average likelihood of a (type of) movie. If you had difficulties with this
question, re-check slide 8 of the ModerationAnalysis rev.pdf file.

Type of movie Model 1 Model 2


Not adventure, romance or comedy 0.055
Pure adventure movie (only adventure, not comedy or romance) 0.055 -.006 =.049
Adventure and romance but not comedy 0.055 -.006 + 0.023 = 0.072
0.095
Comedy and romance but not adventure 0.055 + 0.062 + 0.023 = 0.140
Romance, adventure and comedy 0.055 + 0.062 + 0.023 -.006 =
0.134

In Model 1, there is no interaction effect between rating and movie type. Therefore, Model 1 assumes
that the effect of rating on the likelihood to watch the movie is the same for all types of movies.

On the other hand, Model 2 allows for the effect of rating on likelihood to be different for the different
types of movies through the interaction effects.

2. Answer this question only using the estimates from Model 1. How much more should the rating
of a “pure adventure movie” be higher than a “pure romance movie” in order for movie-goers to
be indifferent in watching both of them (i.e., have the same likelihood) on the same day?

The rating has to be 1.24 points higher for an adventure movie to have the same likelihood as that of a
romance movie.

Likelihood(adventure) = 1.176 + .215 + .095  rating(adventure) -0.005*days

Likelihood(romance) = 1.176 + .333 + .095  rating(romance) – 0.005*days

For Likelihood(adventure) = Likelihood(romance), then

1.176 + .215 + .095  rating(adventure) = 1.176 + .333 + .095  rating(romance)

 rating(adventure) – rating(romance) = (0.333 – 0.215) / 0.095 = 1.24 points higher

You might also like