代写SOC202H1S Introduction to Quantitative Research Methods代做Java语言
- 首页 >> Database作业SOC202H1S
Introduction to Quantitative Research Methods
Due date,submission,and instructions
Please submit an electronic copy on Quercus in a single document that contains all lab components.Please read the detailed instructions about the expected content and formatting of these lab assignments in the document Lab assignment instructions available on Quercus.
Learning Objectives
In this tutorial you will:
● Learn how to create a new variable based on variables in the data set
● Practice formulating hypotheses
● Use SPSS to conduct a two-sample means test
● Use SPSS to graphically compare the means(and confidence intervals)of two groups: Learn how to produce an error bar graph
● Practice interpreting the results of hypothesis tests
Data
The exercise assigned to this week are based on the General Social Survey(GSS).You can download the"GSS2013_shortened.sav"data file from the course website on Quercus(go to "Tutorials">"Datasets").
Exercise 1 (90%of mark)
The research question you will be investigating this week is: Does the average number of friends differ between men and women?
a) For this exercise,you are going to create a new variable capturing respondents'number of friends based on the"Number of Close friends [scf_100c]"and the"Number of other friends [scf_110c]").Generate the new variable by adding two existing variables following the instructions in the tutorial.
Check vour work:Calculate basic descriptive statistics(mean,standard deviation, minimum and maximum).What is the lowest and what is the highest number of friends? What is the average number of friends?Do these values look plausible?
b) Calculate the mean number of friends for women and men.Instruction:Go to Analyze>
Compare Means>Means..>move the variable"Number of friends"into the box "Dependent List"and the variable"Sex of respondent"(sex)into the "Layer 1 of1"box>
click on OK.Include this descriptive table in your report.
● write a sentence of interpretation about how these sample means compare
● explain why we need to conduct a two-sample significance test when we can tell whether there is a difference just by examining the two means in the table
c) Following the instructions in the SPSS DEMONSTRATION 11.2(p.381),produce an
error bar graph comparing women and men's average number of friends.Interpret the graph the point estimates and confidence intervals,separately (one sentence each).Then,interpret both of the graphs together(one sentence only).(That's three sentences in total:One for men,one for women,and one for both of them together)
Next,you are going to conduct a two-sided two-sample test for means at the.05 significance level.
d) State the hypotheses mathematically and literally.
e) Perform. a two-sample means test using Demonstration 11.3(p.382). In order to conduct this hypothesis test,you need to know what the numerical values are that are assigned to the categories "male"and"female"on the variable sex.To find this out,consult the GSS CODE BOOK.
f) State whether you reject or fail to reject the null hypothesis H₀you formulated in part d)at the .05 significance level.
g) Provide a verbal interpretation of the rejection decision.
h) How do the findings from the two-sample means test compare to the findings based on the error bar graph in part e)?(1-2 sentences)
i) What questions do these findings raise?Why might this finding come about?(this is, of course,somewhat speculative.Please limit this to 3 sentences).
Exercise 2 (10%of mark)
Find the variables cerd230c and grp_40.Familiarize yourself with the codebook entries
for both of these variables,especially the missing values.
a)Provide frequency tables for the two variables.If someone states"None"for cerd230c what appears to happen to their response for grp_40?
b)Run a t-test investigating the effect of gender on grp_40 and provide the output before you recode any missing variable.
c)Now run a t-test investigating the effect of gender on grp_40 and provide the
output after you recode any individuals who answer"None"for cerd230c to an interpretable value on grp_40.(see the reading notes for more information about this)
d)What is the difference in interpretation between these two t-tests?Which one seems more useful in this case?(2-3 sentences)
Learning Objectives
In these exercises,you will:
● Practice how to interpret bivariate regression analysis,including the regression slope,the constant, the correlation coefficient rand the coefficient of determination
Data
You may use either the GSS or the Canadian Community Health Survey for this assignment.
Exercise 1(20%of mark)
You may find the SPSS DEMONSTRATION 13.2 at the end of Chapter 13 in the section"You Are the Researcher"(p.447)useful for this section.
Read through the entire directions before you choose variables.Note that you will need to make a causal argument,and one of your interval-ratio variables will be a dependent variable in both analyses.Using either the GSS or Canadian Community Health Survey Dataset,you will select one nominal variable that has more than two categories(non-dichotomous)and two interval-ratio variables.
a) Use the Descriptives command to get means and standard deviation for two interval ratio variables Discuss these descriptive findings in a couple of sentences.(Hint:You did something similar back on Lab A).Include the table in your answer.
b) Use the Frequencies command to get the frequency tables for your nominal variable.Discuss the
distribution of the nominal variable and write a couple of sentences about what the distribution means in real-world terms.(Hint:You did something similar back on Lab A).Include the table in your answer.
c) Transform. your nominal variable into a new dichotomous variable,with"0"representing"No"and"1" representing"Yes."Make sure this new variable has an easy-to-remember name!Then create a two-way frequency table for your old nominal variable and your new nominal variable to make sure the
transformation worked;include the table in your answer.
Exercise 2(40%of mark)
Examining Theory:X=Nominal,Y=Interval/Ratio
a) Generate a causal theory about your nominal variable and your interval/ratio variable.Which one is likely to cause the other,and why?Define xand y.[Note:this should be approximately two sentences;one to
explain which one is likely to cause the other,and one to explain why.If you are having trouble coming up with a relatively plausible theory,you may want to select different variables]
b) Calculate the correlation coefficient.Report the output and interpret the correlation coefficient(one sentence).
c) Run the OLS regression.Report the output,and interpret the regression coefficient and intercept (one sentence for the coefficient,one for the intercept)
d) Interpret R-Squared
e) Test your theory using all six steps from the video lecture and reading notes.When the regression
output gives you what you need,you do not need to calculate it out by hand;simply report what you found from the regression output above.
f) Based on the tests of statistical significance and measures of effect size(correlation coefficient,
regression coefficient,and R-squared),evaluate your theory.Do not write more than 3-4 sentences
Exercise 3(40%of mark)
Examining Theory:X=Interval/Ratio,Y=Interval/Ratio
a) Generate a causal theory about your two interval/ratio variables,using the same dependent variable as in the previous section.Which one is likely to cause the other,and why?Define xand y.[Note:this should be approximately two or maybe three sentences;one to explain which one is likely to cause the other,and one to explain why.If you are having trouble coming up with a relatively plausible theory,you may want to select different variables.]
b) Calculate the correlation coefficient.Report the output and interpret the correlation coefficient(one sentence).
c) Run the OLS regression.Report the output,and interpret the regression coefficient and intercept(ne sentence for the coefficient,one for the intercept)
d) Interpret R-Squared
e) Test your theory using all six steps from the video lecture and reading notes.When the regression
output gives you what you need,you do not need to calculate it out by hand;simply report what you found from the regression output above.
f) Based on the tests of statistical significance and measures of effect size(correlation coefficient, regression coefficient,and R-squared),evaluate your theory.Do not write more than 3or 4 sentences.