Large$-$n assignment $($hypothesis testing with regression techniques$)$

After completing the assignment students should be able to test hypotheses using a given dataset by:

$-$ Constructing scales and providing descriptives of variables $($including but not limited to scale variables$)$

$-$ Testing the model assumptions for multiple linear regression, and using regression to test hypotheses.

Background

A study$$s research objective is to test hypotheses related to individual citizens$$ trust in the municipality of Rotterdam by confronting hypotheses with data gathered in a survey among citizens living in Rotterdam. Scholars have used literature to develop two hypotheses:

$-$ The more individual citizens support a populist ideology, the lower their level of trust in the municipality of Rotterdam will be;

$-$ The more isolated individual citizens are, the lower their level of trust in the municipality of Rotterdam will be$.$

Materials: questionnaire and survey

To measure the variables $$support for populist ideology$,$isolation$$ and $$trust in the municipality of Rotterdam$,$ the following indicators were used in a questionnaire.

VARIABLE Item Question

Supportforpopulistideology POP$1$ What people call compromise in politics is really selling out one$$s principles

POP$2$ Most politicians do not care about people

POP$3$ It is good to have strong political leadership in Rotterdam, even if leadership bends rules if necessary

POP$4$ The municipality of Rotterdam sees its citizens as enemies

Isolation ISO$1$ I feel close to family and friends $($R$)$

ISO$2$ I often feel lonely

ISO$3$ If I feel unhappy, I often turn to others for support $($R$)$

ISO$4$ I often feel there is little meaning in the things I do in daily life

TrustinthemunicipalityofRotterdam TRUST$1$ The municipality of Rotterdam is generally a capable and proficient public service provider

TRUST$2$ I believe the municipality of Rotterdam would do its best to help me if needed

TRUST$3$ The municipality of Rotterdam would keep its commitments

All items were measured using five$-$point Likert items $(1=$strongly disagree, $5=$strongly agree$)$

$($R$)$ indicates reverse$-$coded items.

In the questionnaire, there are also items that ask for a respondent$$s gender $($GENDER $1=$male, $2=$female, $3=$other, $4=$prefer not to disclose$)$ and age $($integer number$).$

The formulations of items are $$given$$ and cannot be changed, and the questionnaire was used in a survey among citizens living in the municipality of Rotterdam.

Materials:

The results of the survey are coded in an SPSS file TrustInRotterdam.SAV. Reverse$-$coded items are re$-$coded $($as ISO$1\_$Reversed and ISO$3\_$Reversed$).$ The variable Gender$\_$Binary is a binary variable with $0=$male $1=$female, and others coded as missing. The data have already been cleaned and unengaged responses have been removed from the data set. The sample size is $200.$

Assignment

Your assignment will be to report the findings of the survey, and more particularly to test the hypotheses $($see above$)$ using this data set. Briefly report the findings. Instructions:

$-$ Briefly report the consistency $($Cronbach$$s alpha$)$ of measurement of variables Support for populist ideology, Isolation and Trust in the municipality of Rotterdam using the items in the questionnaire, and, if possible, construct scale variables using the items $($Supportforpopulistideology$\_$SCALE, Isolation$\_$SCALE, Trustinthe municipalityofRotterdam$\_$SCALE$)$

$-$ Report the descriptives in a single table $($see Table $1,$ add relevant descriptives and statistics for the cells with red borders, ignore grey cells as they are irrelevant or redundant$)$

$-$ Check the model assumptions for multiple linear regression and if possible, use multiple linear regression to test the two hypotheses. Use Table $2$ to report the results of the regression analysis.

$-$ Report $($put in wording, not just in numbers$)$ what statements can be formulated with respect to the hypothesis testing.