Exercises to Accompany the Instructional Subset of 1995 Princeton Survey Research Associates (rc00092) Trust in Government Study
Humboldt State University and
John L. Korey, Calif State Polytechnic University Pomona
© The Authors, 1998; Last Modified 16 August 1998
For these exercises, use data set trussp.por.
TRUS is an instructional subset of the Distrust of Government Study conducted by Princeton Survey Research Associates in 1995. It is consists of 47 variables including measures of trust, confidence, political participation, efficacy, opinion of the media, and numerous demographics.
The exercises, which involve recoding, crosstabulating, and correlation, may be performed by beginning and intermediate students in political science, sociology, psychology, journalism, and communication.
- Are trust and confidence in government related to one's assessment of the nation's economy or one's personal financial situation, or both? Crosstabulate TSTNAT and TSTATE with ECONM and PERSFIN.
- Are trust and confidence in government related to one's trust and confidence generally? Crosstabulate TSTNAT and TSTATE with TSTPEOP, CCORP, CMED, CMEDIA, and CIVIC.
- What is the percentage of the sample who said it would trust both the national government and the state government to do the right thing just about always? Crosstabulate TSTNAT and TSTATE. Include percentage of cells.
- Generate a new variable ETHNIC and label it Ethnicity. Combine the two variables RACE and LATIN so that ETHNIC 1=White, 2=Black, 3=Latin, 4= Asian, and 5=Other.
- Generate a new variable PARSTREN and label it Party Strength. Combine the two variables PARTY and PARLEAN so that PARSTREN 1=Strong Democrat, 2=Moderate Democrat, 3=Leaning Democrat, 4=Independent, 5=Leaning Republican, 6=Moderate Republican, and 7=Strong Republican.
See: Keith, Bruce E. et al. 1992. The Myth of the Independent Voter. Berkeley: The University of California Press.
- Are trust and confidence in Congress and the President related to party identification and ideology? Crosstabulate CCON and CCLINTON with PARSTREN and IDEO.
- Create a correlation (V) matrix of EFF1, EFF2, EFF3, EFF4. Does this matrix confirm or deny the theory of internal and external efficacy?
See: Lane, Robert E. 1959. Political Life: Why People Get Involved in Politics. Glencoe, ILL: The Free Press; Craig, Stephen C. Richard G. Niemei, and Glenn E. Silver. 1990. "Political Efficacy and Trust" Political Behavior 12(3) 189-314, September.
- Generate a new variable EFFINDX and label it Efficacy Index. Combine the four variables EFF1, EFF2, EFF3, and EFF4 to form an efficacy scale of 0 to 12. On each of the four items let a response of Agree Strongly=0, Agree Somewhat=1, Disagree Somewhat=2, and Disagree Strongly=3.
- Are feelings of one's ability to influence politics (EFFINDX) related to one's trust and confidence in government? Is efficacy related to education? Crosstabulate EFFINDX and TSTNAT, TSTATE. Crosstabulate EFFINDX and CNAT. Crosstabulate EFFINDX and EDUC.
- Is voting participation related to efficacy and trust and confidence? Crosstabulate VOT92 and VOT94. Create a new variable, ELECT and label it Electoral Participation with values 0=no participation, 1=participation in one election, and 2=participation in both elections. Crosstabulate ELECT with EFFINDX and with TSTNAT and CNAT.
- Is one's confidence in Congress (CCON) and in the Clinton Administration (CCLINTON) psychological or sociological, related more strongly to one's personality (EFFINDX) or to one's group membership (e.g., ETHNIC, GENDER, PARSTREN).
- Is one's trust and confidence in government related to one's interest in politics and government? Crosstabulate INTEREST with TSTNAT, TSTATE, and CNAT.
- Is one's confidence and trust in government related to one's knowledge of government? Generate a new variable KNOWGOV and label it Knowledge of Government Index. Combine responses to SENATE, MLEADER, VP, SPEAKER, CHIEF, HEALTH, TERMLIM, FAMLEAVE, SPEND to form a scale of 0 to 9.
- What are the characteristics of those who said they trust the national government more than state governments and vice versa? Create new variables TNATMOR and TSTATMOR from the crosstabulation in Exercise 3. Create a new variable CATAGE from AGE and label it Categorical Age. Then crosstabulate TNATMOR and TSTATMORE with GENDER, ETHNIC, CATAGE, EDUC, INCOME, PARSTREN and IDEO.
- Are trust and confidence related to news sources? Is knowledge of government related to news sources? Crosstabulate TSTNAT and TSTATE with NEWS and LIMB. Crosstabulate KNOWGOV with NEWS and LIMB.
See: Miller, Arthur, Edie N. Goldenberg, and Lutz Erbing. 1979. "Type-Set Politics: Impacts of Newspapers on Public Confidence," American Political Science Review. 73:67-84, March.
- Is one's religion related to trust and confidence? Crosstabulate TSTNAT, TSTATE, CNAT with RELIGION and BORNAGN.