Researchers ran a multiple linear regression to determine if belief accuracy in a political claim when given misinformation was related to the following independent variables:
One’s level of anger (coded low = 0, high = 1)
One’s level of anxiety (coded low = 0, high = 1)
Whether the source of misinformation about the claim came from their own political party (in-party, coded from the other party = 0, from their party = 1)
Whether or not a correction of misinformation was made (coded correction without evidence = 0, correction with evidence = 1)
One’s political knowledge (continuous measurement from 0-4)
One of the two issues in question (coded illegal immigration = 0, death penalty = 1)
4. The Adjusted R2 for Model 3 is 0.193. How would one interpret this?
5. In putting together a final model, would you want to remove any variables from Model 3? Why or why not?
6. They did not report standardized regression coefficients (β’s). What would the standardized regression coefficients have told us?
Transcribed Image Text:2:38
◄ Drive
<
The results are listed in the table below:
Effects of Anxiety and Anger,
Partisanship, and Corrections on Belief
Anxiety
Anger
Correction (coded high)
In-party (coded high)
Anxiety x In-party
Anxiety x Correction
Correction In-party
Anxiety x Correction x In-party
Anger x In-party
Anger x Correction
Anger x Correction × In-party
Issue (death penalty coded
high)
Political knowledge
Constant
Observations
F (df)
8+
** p < .01.
*** p < .001.
Model 1
-.14 (11)
-.09 (.11
-.13 (.08
R²
Notes: Unstandardized regression
coefficients reported. Standard error is
listed in parentheses. p-values are
two-tailed.
#p<.10.
*p<.05.
561
.78
.27 (.
.18
1. In Model 1 B-0.015 for political
knowledge How would one interpret
2.18
554)
.15
20.46
:
<>
Expert Solution
Step 1
Answer :
B = 0.015 in Model 1 for political education. How to interpret: Interpret the results of statistical analysis based on the understanding of null hypothesis, P-value, statistics vs. clinical significance, learning power, and evaluation and consideration of statistical errors I and II. B = 0.78 for adjustment in model 1. How can this be explained? Write the regression equation for Model 1. From building a histogram or frequency curve, it can be seen that the data has a normal distribution. However, with a small sample size (n < 20), it may not be clear whether the data from the graph is taken from a normally distributed population. Data may be subjected to formal statistical analysis for evidence of normality, usually using one or more special tests included in computer software packages, such as the Shapiro-Wilks test. Such tests are more reliable with larger sample sizes (n > 100). However, the choice between parametric and nonparametric statistical analysis is less important with a sample of this size, because both analyzes are equally powerful and produce similar results. With a smaller sample size (n < 20), the normality test may be misleading. Unfortunately, nonparametric analysis of small samples does not have statistical power, and it may not be possible to produce a P-value <0.05 despite differences between sample data groups.
