Wednesday, 7 March 2018

Homework 3



1.    Robert Putnam’s research has stimulated interest in the role played by voluntary associations in American democracy. Putnam’s work seems to suggest that, when people get involved in groups and help make collective decisions for the group, they develop participatory skills. These participatory skills, in turn, cause people to participate more in politics – voting at higher rates than people who are not involved in any groups.

A.    This explanation says that the causal relationship between the independent variable and the dependent variable is mediated by an intervening variable.

ANSWER: The independent variable is?
ANSWER: The dependent variable is?
ANSWER: The intervening variable is?

B.    Based on the explanation, write a hypothesis in which the intervening variable is the dependent variable.
 ANSWER:

C.    Based on the explanation, write a hypothesis in which the intervening variable is the independent variable.

ANSWER:

2.    Explain a linear relationship without references to a graph.
ANSWER:

3.    Explain the difference between reliability and validity of measurements. Elaborate. ANSWER:

4.    When is it appropriate to use comparison of means instead of cross tabulation?

ANSWER:

5.    Each of the following conclusions is based on a relationship between X and Y that could be spurious. For each one:
(i)   Identify a plausible control variable, Z.
(ii) Briefly describe how Z might be affecting the relationship between X and Y.

A.    The level of ice cream sales (X) and crime rate (Y) are strongly related: As sales go up, so does the crime rate. Conclusion: To reduce the crime rate, ice cream should be prohibited.
ANSWER: I
ANSWER: II




B.    When one looks at the relationship between marital status (X) and party identification (Y), one finds: Married people are more likely to be Republican than people that are not married. Conclusion: Getting married causes people to become Republican.
ANSWER: I
ANSWER: II

C.    Individuals’ attendance at religious services (X) is related to the number of children they have (Y). Conclusion: Declining religious attendance causes declining birthrates.
ANSWER: I
ANSWER: II


6.    Suppose you want to model a set of interaction relationships between Catholicism, religious attendance, and abortion beliefs. You think that the positive effect of religious attendance on anti-abortion attitudes is significantly stronger for Catholics than non-Catholics. To construct the interaction model, you will build on the base effects of the model
y = a + b1 (Catholic) + b2 (high attendance), where “Catholic” is a Catholic/non-Catholic dummy (Catholics are coded 1, non-Catholics coded 0) and “high attendance” is a high attendance /low attendance dummy (frequent attenders are coded 1, infrequent attenders are coded 0). Before you specify the model, you will need to compute an interaction variable.

Answer:
A.    The interaction variable is computed by multiplying                          times                           .

B.    Which of the following groups of respondents will have a value of 0 on the interaction variable?

i)   Catholic low-attenders
ii)   non-Catholic low-attenders
iii)   Catholic high-attenders
iv)   non-Catholic high-attenders

C.    Which of the following groups of respondents will have a value of 1 on the interaction variable?

i)   Catholic low-attenders
ii)   non-Catholic low-attenders
iii)   Catholic high-attenders
iv)   non-Catholic high-attenders

D.    Write out the interaction model to be estimated.
Answer:






E.    Focus on the coefficient that estimates the interaction effect. If your idea is correctthat the positive effect of religious attendance on anti-abortion attitudes is significantly stronger for Catholics than non-Catholicsthen would you expect the sign on the coefficient to be:

i)   negative
ii)   positive
iii)   close to 0

7.    For this question, please refer to Attachment 1.

A.    Write out the regression equation for this model.
Answer:

B.    Which variable has the strongest relationship with Frequency of contributing money to charity?
          Answer:

C.    How much variation in the dependent variable is explained by the model?
Answer:

D.    If you remove “Age of respondent” from the model, will R2 probably increase, decrease, or stay about the same? Justify your answer.
      Answer:

8.    For this question, please refer to Attachment 2.

A.    Summarize in a sentence the relationship shown in the tables.

Answer:

B.    Identify the control variable.
Answer:

C.    What is the effect of the control variable? (e.g., change in intensity, direction, little or no change). Justify your answer.
Answer:

D.    What does 23.1% in the crosstabs table represent?
Answer:


9.    For this question, a regression model was created and run to test the effect of a respondent’s parents and/or spouse’s level of educational attainment on the respondent’s own level of educational attainment, measured in years of school completed. The output was as follows:

MODEL RESULTS:              R2 (adj.) = .378           F= 109.28        p< .001

Table of Coefficients:

VARIABLE
B
Beta
t
p
(Constant)
5.574



Father’s Highest Yr. School Completed

.095

.136

2.995

.003
Mother’s Highest Yr. School Completed

.104

.121

2.600

.010
Spouse’s Highest Yr. School Completed

.445

.477

12.491

.000

Dependent: Respondent’s Highest Yr. School Completed

Answer the following questions about this model:

A.    Bob’s mom has 12 years of formal education, but his dad found school difficult and dropped out after only 10 years. His wife Margaret, a true braniac, has 20 years of formal education. Make your best prediction of the number of years of formal education Bob has based on the information provided above. How did you arrive at the value of that prediction?
Answer:

B.    Of the independent variables listed, which one has the strongest effect on an individual’s level of educational attainment? Which one has the weakest effect? If you were asked based solely on these variables whether it was more important for someone’s schooling to “be born into brainy” or “marry brainy,” which would you say? What’s the tricky causal issue with that question? Briefly explain your answers.
Answer:

Attachment 1: The following regression model uses data from the General Social Survey to estimate the frequency of contributing money to charities based on socioeconomic status, empathy, degree, and age.
Coefficientsa



Mode


l
Unstandardized Coefficients
Standardized Coefficients


t


Sig.
B
Std. Error
Beta
1
(Constant)
.699
.758


.055
.922
.359

Socioeconomic Index of Respondent
.004
.011
.418
.677

Applicability of description as empathic
.257
.124
.200
2.072
.041

Highest Degree of Respondent
.513
.157
.438
3.273
.002

Age of Respondent
.005
.008
.058
.594
.554
a. Dependent Variable: Frequency of contributing money to charity


Model Summary


Model

R

R Square
Adjusted R Square
Std. Error of the Estimate
1
.514a
.264
.227
1.353
a. Predictors: (Constant), Age of Respondent, Applicability of description as empathic, Socioeconomic Index of Respondent, Highest Degree of Respondent

Attachment 2: Variables Degree (Recoded Degree) and Frequency of giving (recoded Frequency of contributing money to charity).

Frequency of giving * Degree Cross tabulation


Degree


Total
High School or Less
Some College
At least college grad.
frequency
Not
at
all
Count
21
0
0
21
of giving
% within Degree
35.0%
.0%
.0%
23.6%
1-3+ past year
Count
26
4
10
40
% within Degree
43.3%
100.0%
40.0%
44.9%
At least once a
Count
13
0
15
28
month past ye
ar     % within Degree
21.7%
.0%
60.0%
31.5%
Total
Count
60
4
25
89

% within Degree
100.0%
100.0%
100.0%
100.0%


Chi-Square Tests



Value

df
Asymp. Sig. (2-sided)
Pearson Chi-Square
22.578a
4
.000
Likelihood Ratio
28.401
4
.000
Linear-by-Linear Association
16.910
1
.000
N of Valid Cases
89
a. 3 cells (33.3%) have expected count less than 5. The minimum expected count is .94.



Symmetric Measures



Value
Asymp.
Std. Erro a
r
Approx. b
T

Approx. Sig.
Nominal by
Phi
.504


.102


4.809
.000
Nominal
Cramer's V
.356
.000
Ordinal by Ordinal
Gamma
.707
.000
N of Valid Cases

89
a.  Not assuming the null hypothesis.
b.  Using the asymptotic standard error assuming the null hypothesis.


Symmetric Measures


Gender of Respon

dent

Value
Asymp.
Std. Erro a
r
Approx. b
T

Approx. Sig.
MALE
Nominal by
Phi
.592


.148


3.574
.004

Nominal
Cramer's V
.418
.004

Ordinal by Ordinal
Gamma
.670
.000

N of Valid Cases

44
FEMALE
Nominal by
Phi
.487


.129


3.240
.031

Nominal
Cramer's V
.344
.031

Ordinal by Ordinal
Gamma
.785
.001

N of Valid Cases

45
a.  Not assuming the null hypothesis.
b.  Using the asymptotic standard error assuming the null hypothesis.

Chi-Square Tests


Gender of Respondent

Value

df
Asymp. Sig. (2-sided)
MALE
Pearson Chi-Square
15.404a
4
.004

Likelihood Ratio
19.685
4
.001

Linear-by-Linear Association
9.123
1
.003

N of Valid Cases
44
FEMALE
Pearson Chi-Square
10.672b
4
.031

Likelihood Ratio
11.986
4
.017

Linear-by-Linear Association
8.552
1
.003

N of Valid Cases
45
a.  5 cells (55.6%) have expected count less than 5. The minimum expected count is . 82.
b.  6 cells (66.7%) have expected count less than 5. The minimum expected count is . 20.







frequency of giving * Degree * Gender of Respondent Crosstabulation



Gender of Respondent
Degree


Total
High School or Less
Some College
At least college grad.
MALE
frequency of giving
Not at all                 Count
% within Degree
12
46.2%
0
.0%
0
.0%
12
27.3%
1-3+ past year        Count
% within Degree
8
30.8%
3
100.0%
7
46.7%
18
40.9%
At least once a       Count
month past year      % within Degree
6
23.1%
0
.0%
8
53.3%
14
31.8%
Total                                            Count
% within Degree
26
100.0%
3
100.0%
15
100.0%
44
100.0%
FEMALE
frequency of giving
Not at all                 Count
% within Degree
9
26.5%
0
.0%
0
.0%
9
20.0%
1-3+ past year        Count
% within Degree
18
52.9%
1
100.0%
3
30.0%
22
48.9%
At least once a       Count
month past year      % within Degree
7
20.6%
0
.0%
7
70.0%
14
31.1%
Total                                            Count
% within Degree
34
100.0%
1
100.0%
10
100.0%
45
100.0%

No comments:

Post a Comment

Balsa Glider Project Assignment

Balsa Glider Project Assignment #2 An APA paper must be submitted which includes the following sections (with headings): Prototype...