POLI 300                            PROBLEM SET #3A                                             10/03/05


VARIABLES: ANSWERS AND DISCUSSION



Note: These are sample answers; other names for some variables may be equally appropriate. In some cases, different variables pertaining to different units of analysis may also be appropriate.



 

             VARIABLE 1                                                  VARIABLE 2                    units [pop.]

          (range of values)                                          (range of values)


 

1.   LEVEL OF SENIORITY                     DEGREE OF PRAGMATISM            individuals

       (LO-HI, or years/terms)                           (LO-HI, or index score)            [members of Congress]


 

2.   LEVEL OF EDUCATION                  DEGREE OF RELIGIOSITY             individuals

          (LO-HI, or years)                           (LO-HI, indicator or index score)


 

3. WHETHER [or RATE AT WHICH]         MURDER RATE                               legal jurisdictions

      CAPITAL PUNISHMENT IMPOSED  (LO-HI, or numerical rate)                   [e.g., states]

      (YES/NO [LO-HI, or numerical rate])



4. DEGREE OF COMPETITIVENESS   DEGREE OF RESPONSIVENESS   Congressmen and

             (LO-HI, or index score)                       (LO-HI, or index score)                their districts*


 

5.   LEVEL OF PRESIDENT'S                     LEVEL OF PRESIDENT'S              elections [Pres.

      APPROVAL RATING                             RE-ELECTION VOTE                    elections with

            (LO-HI, or %)                                       (LO-HI, or %)                            incumbent running]

 

6. AMOUNT OF STUDY EFFORT          LEVEL OF GRADES/GPA                individuals

         (LO-HI, or hours/week)                   (LO-HI, F-A, or numerical GPA)          [students]


 

7. DEGREE OF CLOSENESS               LEVEL OF VOTING TURNOUT        elections

           (LO-HI or % dif)                                     (LO-HI or %)                           (longitudinal)

                                                               or

DEGREE OF CLOSENESS           LEVEL OF VOTING TURNOUT      Cong.(etc.) districts

            (LO-HI or % dif)                                    (LO-HI or %)                           (cross-sectional)


8.   DEGREE OF GOODNESS OF TIMES     PER CENT OF INCUMBENT          elections

        (L0-HI, or index score)                              CANDIDATES REELECTED 

                                                                              (0%-100%)

 

9.   GOODNESS OF SLEEP HABITS           LEVEL OF WELL-BEING                individuals

                         (LO-HI)                          (LO-HI, or INCOME as indicator)


 

10. WHETHER OR NOT EAT AN            NUMBER OF DOCTOR VISITS           individuals

          APPLE A DAY (YES/NO)                          (LO-HI, count per year)

                                                            or (less literally)

      GOODNESS OF EATING HABITS        LEVEL OF HEALTH                        individuals

                        (LO, HI)                                             (LO, HI)


 

11. AMOUNT OF EDUCATION              LEVEL OF SUCCESS                             individuals

        (LO-HI, or number of years)                (LO-HI, or INCOME)

                                                   [Note: not LEVEL OF DESIRE TO SUCCEED]


 

12.  IDEOLOGY OF GOVERNMENT        LEVEL OF INFLATION                nations

                (LEFT-RIGHT)                                  (LO-HI or %/year)


 

13. LEVEL OF POLITICAL INTEREST    WHETHER OR NOT VOTES          individuals

                         (LO-HI)                                  (YES, NO)                              [eligible voters]


 

14.  DEGREE OF LIBERALISM                  DIRECTION OF VOTING             individuals

       (LO, HI) [or IDEOLOGICAL                 (DEM, IND, REP)                         [U.S. voters]

       PLACEMENT (LIB-CONS)]


 

15. LEVEL OF SATISFACTION             VOTE CHOICE                                 individuals

       WITH ECONOMY (LO-HI)           (INCUMBENT, CHALLENGER)        [voters]


 

16.  TYPE OF ELECTORAL SYSTEM       NUMBER OF PARTIES                 nations [with

   (MAJORITARIAN, PROPORTIONAL)      (2, more than 2)                           free elections]

Note: Statement #16 is “Duverger's Law,” perhaps the most famous law-like generalization in political science.


17.  AVERAGE STRENGTH OF PARTY              TIME                                             electorates (in

              IDENTIFICATION                          EARLIER, LATER [1972 – 2004]         successive

       (LO-HI, or numerical average)                                                                                  elections)

Note: Statement #17 makes a longitudinal statement (claiming a pattern over time), so the second variable is TIME. The range of values might be from 1972 through 2004 (e.g., as in the SETUPS data).**


 

18.                SIZE                                 LEVEL OF EFFECTIVENESS                            legislatures

       (LO-HI or number of members)                (LO-HI or index score)


 

19.  WHETHER OR NOT TERM-            LEVEL OF EFFECTIVENESS                        legislatures

             LIMITED (YES, NO)                      (LO-HI or index score)


Note 1. In general, students seem to have an easier time identifying the variables than the unit of analysis, especially when the latter is anything other than individuals.  Remember that the variables refers to characteristics of the “things” that constitute the unit of analysis.  For example, it does not make sense to say that individuals (voters) constitute the unit of analysis in #7 and also that DEGREE OF CLOSENESS is one of the variables, since voters do not vary in closeness (rather, elections do). 

More generally, imagine that you set out to collect data that bears on the empirical proposition implicit in each sentence.  This means that you “observe” or “measure” the value of each variable for each case you study.  Thus for each sentence, you would have two columns of data when you enter the data into a spreadsheet similar the Student Survey data that was returned to you.  Each pair of entries in each column would records the value of the two variables in a particular case and each row of the data sheet corresponds to a case. Think about what kind of thing each case/row represents.  Are they individuals, nations, elections, legislatures, or what?  Whatever they are, they define the unit of analysis. 

Note 2.  The range of possible values for a variable is best summarized as “low to high” (or whatever), rather than either “low or high” or “low and high” (or whatever), since the latter expressions suggests the variable is intrinsically dichotomous (two-valued), which typically is not true — that is, there are typically several or many intermediate values between the two extremes.  More generally, don’t “dichotomize” variables unnecessarily,  e.g., DEGREE OF CLOSENESS (of elections) is preferable to WHETHER/NOT CLOSE.  Moreover, the range of possible values needs to be specified separately for each variable (since they may be different).

*    Since (at any given time) there is a one-to-one correspondence between House members and districts, either can be considered the unit of analysis.