Answering Social Research Questions with Statistical Models (SOST20131&30031), 2022-23

Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

Answering Social Research Questions with Statistical Models (SOST20131&30031) , 2022- 23

Assessment, Part 1 2022

Submission deadline: 10 AM, Friday 2nd December 2022, via Blackboard/turnitin

Word limit: 1200 words                                                                                                                                        The research seeks to answer the following research question:

“What is the total causal effect of being unemployed on the level of life satisfaction?”                         You are considering several factors that are important determinants of life satisfaction. Drawing       upon relevant literature and existing theories, you hypothesise the causal framework as depicted     below. Examine the causal pathways in DAG.

1. DAG representing unemployment and life satisfaction

Marriage

Sex

Life satisfaction

Figure 1: DAG representing unemployment and life satisfaction

The mean level of life satisfaction was 6.3 (SD = 2.01). The participants were 44 years of age on average (SD=12.26, ranged from age 25 to 65).

• Health status (i.e. good or poor health) is a common cause for unemployment and life satisfaction.

• Being jobless causes home ownership status, which then affects the level individuals feel satisfied with their life.

• The prospects for marriage is affected by the employment status, which in turn, causes the extent to which individuals are satisfied with their life.

• Unemployment is affected by age, and life satisfaction is affected by age, as well.

• The educational attainment level has a causal relationship with both employment      status and life satisfaction, affecting both. Education is regarded as a common cause.

• Gender affects the level of life satisfaction.

• You are primarily interested in these variables shown above, and assume the rest of the individual circumstances are the same, to pursue the research.

2. Model results

Table 1 reports several multiple linear regression models.

Table 1: Results (β coefficients and R-squares) from the five multiple linear regression models predicting life satisfaction.

Model 1

β

Model 2

β

Model 3

β

Model 4

β

Model 5

β

Unemployed

Age (in years)

Women

Marriage (Have a spouse) Home ownership Degree holder

Poor health

-0.75**

-0.62** 0.08** 0.40*** 0.70** 0.24***

-0.58**

-0.09**

0.12**

-0.22*

-0.52**

-0.07**

0.68**

0.21***

0.16

-0.20*

-0.52**

-0.05**

0.40***

0.63**

0.20***

0.12*

-0.18*

Intercept

6.28

6.83

7.55

7.49

8.12

R2

0.15

0.16

0.26

0.29

0.20

Note: N = 1,167 for all models. Reference category: In employment, men, no spouse, not owning        home, highest educational attainment is below degree-level and in good health. Age is a continuous  measure. The rest of variables are dummy variables. Significance level: *p<0.1; **p<0.05; ***p<0.01.

Questions:

1.    Identify the treatment variable. [10 marks]

2.    Using the causal relationships shown in the DAG in Figure 1, say which model(s) of those      shown in Table 1 you would select to answer your research question. Using the appropriate terminology in DAG, fully explain and justify your answer. [30 marks]

3.    For the model(s) you selected in question 2 above, write out the equation. Interpret the       model fit in relation to the less complex model. Then, interpret the chosen model using all   coefficients in the model results. Provide a comparison with other model(s) where possible. In doing so, make sure that you give a clear answer to the research question. [30 marks]

4. Using the model 5 results, provide the calculation of the model-predicted value of the life satisfaction for the following type of participant:

a. A 50-year old unemployed married (has a spouse) man, who does not own home, who is a degree-holder, and who is in good health [10 marks]

b. A 35-year old employed woman, who has no spouse, who owns home, who does not hold a degree, and who is in poor health [10 marks]

5.    Make a comment on Table 1. Does it include essential information to assess the regression models? Provide reasons. [10 marks]

Guidance notes

•    Your submission should answer each of the 5 questions above.

•    Good answers are those that clearly address all parts of the question. Supporting and             justifying your answers with references to books and articles in the course reading lists, and  other scholarly material, is encouraged and will help contribute to a higher mark. Provide a   full reference for any work that you do cite, using the Harvard system of referencing. Include no more than 5 citations in your answer.

•   The word limit for your submission is 1,200 words. Note that this is a limit, not a target; you are permitted to use fewer than 1,200 words, but not more.

•    Your answers to questions 2 and 3 will be marked without regard to whether you answered question 2 correctly, i.e. you will not gain or lose marks for questions 2 and 3 based on your answer to question 2.


2022-11-28