2.32      6.61      6.90      8.04      9.45
10.26     11.34     11.63     12.66     12.95
13.67     13.72     14.35     14.52     14.55
15.01     15.33     16.55     17.15     18.22
18.30     18.71     19.54     19.55     20.58
20.89     20.91     21.13     23.85     26.04
27.07     28.76     29.15     30.54     31.99
32.82     33.26     33.80     34.76     36.22
37.52     39.28     40.80     43.97     45.58
52.36     61.57     63.85     64.30     69.49

Make a histogram (bar graph) of these data using intervals of $5 (i.e., $0.00-$4.99, $5.00-$9.99, $10.00-$14.99, etc.). Make sure the bar height of your histogram is proportional to the category frequency (i.e., the number of observations in each category). Write down the category frequencies at the top of their corresponding bars. Describe the distributions and provide appropriate descriptive statistics.  (Descriptive statistics are discussed in chapter 2 of Wonnacott and Wonnacott.)

II. Babe Ruth's 15 yearly home run totals were

54  59  35  41  46  25  47  60  54  46  49  46  41  34  22

Describe the distributions and provide appropriate descriptive statistics.
 

III. Analyze the 1984 Gallup Survey data.  The data have been saved in an SPSS system file with the name GALLUP.SAV.  The relevant variable names are given in the code book handed out with this lab.  In order to complete the lab, you will need to read Aldrich et al., "Foreign Affairs and Issue Voting."  Don't worry about the parts of the article which are beyond the scope of this course.  

(1) Attempt to replicate the results reported in Table 2 (p129) and the 1984 Gallup part of Tables A-1 and A-2 (p137-138) of the article. Explain, if you can, the discrepancy between your results and those of the article. Discuss, briefly, what conclusions Aldrich et al. are trying to draw from these tables. (Note: Follow the footnotes of Tables A-1 and A-2 and, when appropriate, reverse scales so that 1 = most liberal and 7 = most conservative.)

 The SPSS syntax notes which follow should be of great help in replicating the results.
 

Note:  for replication of Table 2,
 missing values v056 v063 (8,9)l.
 frequencies variables=v056 v063.

Note:  for replication of "Strong or Tough Stance" in Tables A-1, A-2,
 missing values v120 to v122 (8,9).
 recode v120 to v122 (1=7) (2=6) (3=5) (5=3) (6=2) (7=1).
 if (v120 <= 7) v120a=1.
 if ((v120 <= 7) and (v121 <= 7) and (v122 <= 7)) v120b=1.
 if ((v120b = 1) and (v121 <> v122)) v120c=1.
 if ((v120b = 1) and (v121 > v122)) v120d=1.
 frequencies variables=v120a v120b v120c v120d.
 frequencies variables=v120 v121 v122 /statistics=median.

(2) Construct, using information given in both the article and the code book, the independent variables (i.e., "International," "Domestic," "Party Identification," and "Candidate Evaluation") used in the probit estimated vote equations reported in Table 5 (for the 1984 Gallup Data part only) (p134). Provide descriptive statistics for these variables by treating them as interval scales. Keep in mind that you are not replicating the results of Table 5, but are simply providing descriptive statistics for the independent variables used in this table.  Remember that you need to define missing values and reverse scales whenever appropriate.

Note:  define missing values and reverse scales if necessary
 missing values _____________.
 recode _____________________.

Note:  constructing "International" (iscale) 
 compute iscale=v___+v___+v___+v___+v___+v___. 

Note:  constructing "Domestic" (dscale)
 compute dscale=v___+v___+v___+v___+v___+v___. 

Note:  constructing "Party Identification" (pid)
 if ((v023 = 2) and (v023a = 1)) pid=1. 
 if ((v023 = 2) and (v023a = 2)) pid=2. 
 if ((v023 = 3) and (v023b = 1)) pid=3. 
 if ((v023 = 3) and (v023b = 3)) pid=4. 
 if ((v023 = 3) and (v023b = 2)) pid=5. 
 if ((v023 = 1) and (v023a = 2)) pid=6. 
 if ((v023 = 1) and (v023a = 1)) pid=7. 

Note:  constructing "Candidate Evaluation" (rmlike)
 recode v146 (1=7)(2=6)(3=5)(5=3)(6=2)(7=1). 
 recode v152 (1=7)(2=6)(3=5)(5=3)(6=2)(7=1). 
 compute rmlike=v146-v152. 

Note:  computing summary statistics for interval variables 
 descriptives iscale dscale pid rmlike
   /statistics=min max mean stddev variance.