Analysing the American National Election Study of 1948 with memisc

Introduction

This vignette gives an example for the analysis of a typical social science data set. It is the data file of the American National Election Study of 1948¹ available from the American National Election Studies website. The data file contains data from to USA-wide surveys conducted October and November 1948 by the Survey Research Centre, University Michigan (principal investigators: Angus Campbell and Robert L. Kahn). The total number of cases in the data set is 662 and the number of variables is 65 (more details about this data set can be found at https://electionstudies.org/studypages/1948prepost/1948prepost.htm).

With 662 cases and 65 variables, the 1948 ANES data set is relatively small as compared to current social science data sets. Such larger data sets can be processed along the same lines as in this vignette. Unlike the 1948 ANES data, their size as well as, in some cases, legal restrictions prohibit the inclusion of such a data set into the package, however.

This vignette starts with a demonstration how a data file can be examined before loading it and how a subset of the data can be loaded into memory. After loading this subset into memory, some descriptive analyses are conducted that showcase the construction of contingency tables and of general tables of descriptive statistics using the genTable function. In addition, a logit analysis is demonstrated and the collection of several logit coefficients into a comprehensive table by the mtable function.

It should be noted that the analyses reported in the following are conducted only for purpose of demonstrating the features of the package and are not to be considered of conclusive scientific evidence of any kind.

This vignette is run with the help of the knitr package. This allows to showcase not only data management facilities provided by memisc. The following code also demonstrates how output created with some of the facilities of memisc can neatly integrated in reports generated with knitr. Before we start, we adjust knitr’s output (with which this vignette is formatted) to produce HTML where possible.

knit_print.codebook <-function(x,...) 
  knitr::asis_output(format_html(x,...))

knit_print.descriptions <-function(x,...) 
  knitr::asis_output(format_html(x,...))

knit_print.ftable <-function(x,options,...)
  knitr::asis_output(
    format_html(x,
                digits=if(length(options$ftable.digits))
                          options$ftable.digits
                       else 0,
                ...))
# We can now adjust the number of digits after the comma
# for each column e.g. by adding an `ftable.digits` option
# to an R chunk, as in ```{r,ftable=c(2,2,0)}

knit_print.mtable <-function(x,...)
  knitr::asis_output(format_html(x,...))

Reading in a “portable” SPSS data file

We start with importing the data into R. The following code extracts the SPSS portable file NES1948.POR from zip file NES1948.ZIP delivered with the memisc package.

library(memisc)
options(digits=3)
nes1948.por <- unzip(system.file("anes/NES1948.ZIP",package="memisc"),
                     "NES1948.POR",exdir=tempfile())

Now the portable file is in a temporary directory and the path to the file is contained in the string variable nes1948.por. In the next step, the file is declared as a SPSS/PSPP portable file using the function spss.portable.file, which as first argument takes the path to the file. spss.portable.file reads in the information about the variables contained in the data set and counts the number of cases in the file. That is, standard I/O operations are used on the file, but the data read in are just thrown away without allocating core memory for the data. This counting of cases can, of course, be suppressed if it would take too long.

nes1948 <- spss.portable.file(nes1948.por)

Warning: 9 variables have duplicated labels:
  V480004, V480012, V480020, V480021A, V480021B, V480033A, V480033B,
  V480034A, V480034B

print(nes1948)


SPSS portable file '/tmp/Rtmp6Elp9n/file67051ffcca09/NES1948.POR' 
    with 67 variables and 662 observations

At this stage, the data are not loaded into the memory yet. But we can see which variables exist inside the data set:

names(nes1948)

 [1] "VVERSION" "VDSETNO"  "V480001"  "V480002"  "V480003"  "V480004" 
 [7] "V480005"  "V480006"  "V480007"  "V480008"  "V480009"  "V480010" 
[13] "V480011"  "V480012"  "V480013"  "V480014A" "V480014B" "V480015A"
[19] "V480015B" "V480016A" "V480016B" "V480017A" "V480017B" "V480018" 
[25] "V480019"  "V480020"  "V480021A" "V480021B" "V480022A" "V480022B"
[31] "V480023"  "V480024"  "V480025A" "V480025B" "V480026"  "V480027" 
[37] "V480028"  "V480029"  "V480030"  "V480031A" "V480031B" "V480031C"
[43] "V480032A" "V480032B" "V480032C" "V480033A" "V480033B" "V480034A"
[49] "V480034B" "V480035A" "V480035B" "V480036A" "V480036B" "V480037" 
[55] "V480038"  "V480039"  "V480040"  "V480041"  "V480042"  "V480043" 
[61] "V480044"  "V480045"  "V480046"  "V480047"  "V480048"  "V480049" 
[67] "V480050"

Note that the variable names are all changed from uppercase to lowercase (SPSS does not distinguish uppercase and lowercase variable names and uppercase looks like shouting). Casefolding could have been suppressed by the call spsp.portable.file(nes1948.por,tolower=FALSE).

We also can ask for a description (“variable label”) for each variable:

description(nes1948)

Variable	Description
VVERSION	NES VERSION NUMBER
VDSETNO	NES DATASET NUMBER
V480001	ICPSR ARCHIVE NUMBER
V480002	INTERVIEW NUMBER
V480003	POP CLASSIFICATION
V480004	CODER
V480005	NUMBER OF CALLS TO R
V480006	R REMEMBER PREVIOUS INT
V480007	INTR INTERVIEW THIS R
V480008	PRVS PRE-ELCTN R REINT
V480009	R INT IN PRE/POSTELCTN
V480010	RENT CNTRL KEPT/DROPPED
V480011	GOVT CONTROL PRICES
V480012	WHAT TO DO W TFT-HT ACT
V480013	PRESLELCTN OTCM SURPRISE
V480014A	WHY PPL VTD FOR TRUMAN 1
V480014B	WHY PPL VTD FOR TRUMAN 2
V480015A	WHY PPL VTD AGNST TRUMAN 1
V480015B	WHY PPL VTD AGNST TRUMAN 2
V480016A	WHY PPL VTD FOR DEWEY 1
V480016B	WHY PPL VTD FOR DEWEY 2
V480017A	WHY PPL VTD AGNST DEWEY 1
V480017B	WHY PPL VTD AGNST DEWEY 2
V480018	DID R VOTE/FOR WHOM
V480019	WN DECIDE FOR WHOM TO VT
V480020	CNSD VT FOR SOMEONE ELSE
V480021A	XWHY DID NOT VT FOR HIM 1
V480021B	XWHY DID NOT VT FOR HIM 2
V480022A	WHY VT THE WAY YOU DID 1
V480022B	WHY VT THE WAY YOU DID 2
V480023	VOTED STRAIGHT TICKET
V480024	R NOT VT-IF VT,FOR WHOM
V480025A	R NOT VT-WHY DID NOT VT 1
V480025B	R NOT VT-WHY DID NOT VT 2
V480026	R NOT VT-WAS R REG TO VT
V480027	VTD IN PRVS PRESL ELCTN
V480028	VTD FOR WHOM IN 1944
V480029	OCCUPATION OF HEAD
V480030	HEAD BELONG TO LBR UN
V480031A	GRPS IDENTIFIED W TRUMAN 1
V480031B	GRPS IDENTIFIED W TRUMAN 2
V480031C	GRPS IDENTIFIED W TRUMAN 3
V480032A	GRPS IDENTIFIED W DEWEY 1
V480032B	GRPS IDENTIFIED W DEWEY 2
V480032C	GRPS IDENTIFIED W DEWEY 3
V480033A	ISSUES CONNECTED W TRMN 1
V480033B	ISSUES CONNECTED W TRMN 2
V480034A	ISSUES CONNECTED W DEWEY 1
V480034B	ISSUES CONNECTED W DEWEY 2
V480035A	PERSONAL ATTRIBUTE TRMN 1
V480035B	PERSONAL ATTRIBUTE TRMN 2
V480036A	PERSONAL ATTRIBUTE DEWEY 1
V480036B	PERSONAL ATTRIBUTE DEWEY 2
V480037	CMPN INCIDENTS MENTIONED
V480038	41-PRESLELCTN PLAN TO VT
V480039	41-PLAN TO VT REP/DEM
V480040	41-USA’S CNCRN W OTHERS
V480041	41-SATISD USA TWRD RUSS
V480042	41-INFORMATION LEVEL
V480043	41-USA GV IN,AGRT RUSS
V480044	41-USA-RUSS AGRT VIA U.N
V480045	SEX OF RESPONDENT
V480046	RACE OF RESPONDENT
V480047	AGE OF RESPONDENT
V480048	EDUCATION OF RESPONDENT
V480049	TOTAL 1948 INCOME
V480050	RELIGIOUS PREFERENCE

or even a code book using

codebook(nes1948)

(this is not shown here because the output would have taken more then thirty pages). We can also get a codebook of the first few variabels instead, with

codebook(nes1948[1:5])

VVERSION — ‘NES VERSION NUMBER’

Storage mode:	double
Measurement:	interval

Min:	1	.
Max:	1	.
Mean:	1	.
Std.Dev.:	0	.

VDSETNO — ‘NES DATASET NUMBER’

Storage mode:	character
Measurement:	nominal

Min:	"1948		T"
Max:	"1948		T"

V480001 — ‘ICPSR ARCHIVE NUMBER’

Storage mode:	double
Measurement:	interval

Min:	7218	.
Max:	7218	.
Mean:	7218	.
Std.Dev.:	0	.

V480002 — ‘INTERVIEW NUMBER’

Storage mode:	double
Measurement:	interval

Min:	1001	.	000
Max:	1662	.	000
Mean:	1331	.	500
Std.Dev.:	191	.	103

V480003 — ‘POP CLASSIFICATION’

Storage mode:	double
Measurement:	nominal

Values and labels		N	Percent
1	‘METROPOLITAN AREA’	182	27	.	5
2	‘TOWN OR CITY’	354	53	.	5
3	‘OPEN COUNTRY’	126	19	.	0

Reading in a subset of the data

After we have decided which variables to use we can read in a subset of the data:

vote.48 <- subset(nes1948,
              select=c(
                  V480018,
                  V480029,
                  V480030,
                  V480045,
                  V480046,
                  V480047,
                  V480048,
                  V480049,
                  V480050
                  ))

The subset of the ANES 1948 we read in is now contained in the variable vote.48, which contains an object of class data.set. A data.set is an “embellished” version of a data.frame, a data structure intended to contained labelled vectors. labelled vectors contain the all the special information attached to the variables in the original data set, such as variable labels, value labels, and general missing values. A short summary of this special information shows up after a call to str.

str(vote.48)

Data set with 662 obs. of 9 variables:
 $ V480018: Nmnl. item w/ 7 labels for 1,2,3,... + ms.v.  num  1 2 1 2 1 2 2 1 2 1 ...
 $ V480029: Nmnl. item w/ 12 labels for 10,20,30,... + ms.v.  num  70 30 40 10 10 20 80 80 40 40 ...
 $ V480030: Nmnl. item w/ 4 labels for 1,2,8,... + ms.v.  num  1 2 2 2 2 2 2 2 1 1 ...
 $ V480045: Nmnl. item w/ 3 labels for 1,2,9 + ms.v.  num  1 2 2 2 1 2 1 2 1 1 ...
 $ V480046: Nmnl. item w/ 4 labels for 1,2,3,... + ms.v.  num  1 1 1 1 1 1 1 1 1 1 ...
 $ V480047: Nmnl. item w/ 7 labels for 1,2,3,... + ms.v.  num  3 3 2 3 2 3 4 5 2 2 ...
 $ V480048: Nmnl. item w/ 4 labels for 1,2,3,... + ms.v.  num  1 2 2 3 3 2 1 1 2 2 ...
 $ V480049: Nmnl. item w/ 8 labels for 1,2,3,... + ms.v.  num  4 7 5 7 5 7 5 2 5 6 ...
 $ V480050: Nmnl. item w/ 6 labels for 1,2,3,... + ms.v.  num  1 1 2 1 2 1 1 1 1 2 ...

This output shows, for example, that variable V480018 has the description (variable label) “DID R VOTE/FOR WHOM” is considered as having nominal level of measurement, has seven value labels and one defined missing value.

Since the variable names in the ANES data set are not very mnemonic, we rename the variables:

vote.48 <- rename(vote.48,
                  V480018 = "vote",
                  V480029 = "occupation.hh",
                  V480030 = "unionized.hh",
                  V480045 = "gender",
                  V480046 = "race",
                  V480047 = "age",
                  V480048 = "education",
                  V480049 = "total.income",
                  V480050 = "religious.pref"
        )

Since many data sets available from public repositories have such non-mnemonic variable names as in this example, it might be convenient to do the data loading and renaming in one step. Indeed it is possible:

vote.48 <- subset(nes1948,
                  select=c(
                    vote           = V480018,
                    occupation.hh  = V480029,
                    unionized.hh   = V480030,
                    gender         = V480045,
                    race           = V480046,
                    age            = V480047,
                    education      = V480048,
                    total.income   = V480049,
                    religious.pref = V480050
                  ))

Before we start with analyses, we take a closer look at the data.

codebook(vote.48)

vote — ‘DID R VOTE/FOR WHOM’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘VOTED - FOR TRUMAN’	212	32	.	1	32	.	0
2		‘VOTED - FOR DEWEY’	178	27	.	0	26	.	9
3		‘VOTED - FOR WALLACE’	1	0	.	2	0	.	2
4		‘VOTED - FOR OTHER’	11	1	.	7	1	.	7
5		‘VOTED - NA FOR WHOM’	20	3	.	0	3	.	0
6		‘DID NOT VOTE’	238	36	.	1	36	.	0
9	M	‘NA WHETHER VOTED’	2				0	.	3

occupation.hh — ‘OCCUPATION OF HEAD’

Storage mode:	double
Measurement:	nominal
Missing values:	99

Values and labels			N	Valid			Total
10		‘PROFESSIONAL, SEMI-PROFESSIONAL’	44	6	.	9	6	.	6
20		‘SELF-EMPLOYED, MANAGERIAL, SUPERVISORY’	73	11	.	5	11	.	0
30		‘OTHER WHITE-COLLAR (CLERICAL, SALES, ET’	79	12	.	5	11	.	9
40		‘SKILLED AND SEMI-SKILLED’	164	25	.	9	24	.	8
60		‘PROTECTIVE SERVICE’	6	0	.	9	0	.	9
70		‘UNSKILLED, INCLUDING FARM AND SERVICE W’	85	13	.	4	12	.	8
80		‘FARM OPERATORS AND MANAGERS’	105	16	.	6	15	.	9
92		‘STUDENT’	7	1	.	1	1	.	1
94		‘UNEMPLOYED’	5	0	.	8	0	.	8
95		‘RETIRED, TOO OLD OR UNABLE TO WORK’	38	6	.	0	5	.	7
96		‘HOUSEWIFE’	28	4	.	4	4	.	2
99	M	‘NA’	28				4	.	2

unionized.hh — ‘HEAD BELONG TO LBR UN’

Storage mode:	double
Measurement:	nominal
Missing values:	8 - Inf

Values and labels			N	Valid			Total
1		‘YES’	150	23	.	3	22	.	7
2		‘NO’	493	76	.	7	74	.	5
8	M	‘DK’	5				0	.	8
9	M	‘NA’	14				2	.	1

gender — ‘SEX OF RESPONDENT’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘MALE’	302	45	.	8	45	.	6
2		‘FEMALE’	357	54	.	2	53	.	9
9	M	‘NA’	3				0	.	5

race — ‘RACE OF RESPONDENT’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘WHITE’	585	90	.	7	88	.	4
2		‘NEGRO’	60	9	.	3	9	.	1
3		‘OTHER’	0	0	.	0	0	.	0
9	M	‘NA’	17				2	.	6

age — ‘AGE OF RESPONDENT’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘18-24’	57	8	.	7	8	.	6
2		‘25-34’	142	21	.	7	21	.	5
3		‘35-44’	174	26	.	6	26	.	3
4		‘45-54’	125	19	.	1	18	.	9
5		‘55-64’	86	13	.	1	13	.	0
6		‘65 AND OVER’	70	10	.	7	10	.	6
9	M	‘NA’	8				1	.	2

education — ‘EDUCATION OF RESPONDENT’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘GRADE SCHOOL’	292	44	.	4	44	.	1
2		‘HIGH SCHOOL’	266	40	.	4	40	.	2
3		‘COLLEGE’	100	15	.	2	15	.	1
9	M	‘NA’	4				0	.	6

total.income — ‘TOTAL 1948 INCOME’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘UNDER $500’	25	3	.	8	3	.	8
2		‘$500-$999’	43	6	.	6	6	.	5
3		‘$1000-1999’	110	16	.	8	16	.	6
4		‘$2000-2999’	185	28	.	2	27	.	9
5		‘$3000-3999’	142	21	.	7	21	.	5
6		‘$4000-4999’	66	10	.	1	10	.	0
7		‘$5000 AND OVER’	84	12	.	8	12	.	7
9	M	‘NA’	7				1	.	1

religious.pref — ‘RELIGIOUS PREFERENCE’

Storage mode:	double
Measurement:	nominal
Missing values:	9

Values and labels			N	Valid			Total
1		‘PROTESTANT’	460	70	.	0	69	.	5
2		‘CATHOLIC’	140	21	.	3	21	.	1
3		‘JEWISH’	25	3	.	8	3	.	8
4		‘OTHER’	14	2	.	1	2	.	1
5		‘NONE’	18	2	.	7	2	.	7
9	M	‘NA’	5				0	.	8

We now have obtained a codebook, which contains information of the class and type of the variables in the data set, the value labels and defined missing values, and counts of the distinct values of the variables.

Analysis

Some descriptive analyses

We start our analyses with a contingency table, but first we make some preparations: We recode the variables of interest into a smaller number of categories in order to get results that are easier to read and interpret.

vote.48 <- within(vote.48,{
  vote3 <- recode(vote,
    1 -> "Truman",
    2 -> "Dewey",
    3:4 -> "Other"
    )
  occup4 <- recode(occupation.hh,
    10:20 -> "Upper white collar",
    30 -> "Other white collar",
    40:70 -> "Blue collar",
    80 -> "Farmer"
    )
  relig3 <- recode(religious.pref,
    1 -> "Protestant",
    2 -> "Catholic",
    3:5 -> "Other,none"
    )
   race2 <- recode(race,
    1 -> "White",
    2 -> "Black"
    )
  })

Warning in recode(vote, "Truman" <- 1, "Dewey" <- 2, "Other" <- 3:4): recoding
created 260 NAs

Warning in recode(occupation.hh, "Upper white collar" <- 10:20, "Other white
collar" <- 30, : recoding created 106 NAs

Warning in recode(religious.pref, "Protestant" <- 1, "Catholic" <- 2,
"Other,none" <- 3:5): recoding created 5 NAs

Warning in recode(race, "White" <- 1, "Black" <- 2): recoding created 17 NAs

Having constructed the unordered factors vote3, occup4, relig3, and race2 we can proceed examining the association the vote, occupational class, relgious denomination, and race. First, we look upon a simple contingency table.

ftable(xtabs(~vote3+occup4,data=vote.48))

	occup4
vote3	Upper white collar	Other white collar	Blue collar	Farmer
Truman	17	30	114	26
Dewey	67	31	36	14
Other	2	0	4	3

Tables of percentages may seem more informative about the impact of various factors on the vote. So we use the function genTable to obtain such tables of percentages:

gt1 <- genTable(percent(vote3)~occup4,data=vote.48)
## For knitr-ing, we use ```{r, ftable.digits=c(2,2,2,0)} here.
ftable(gt1,row.vars=2)

occup4	Truman			Dewey			Other			N
Upper white collar	19	.	77	77	.	91	2	.	33	86
Other white collar	49	.	18	50	.	82	0	.	00	61
Blue collar	74	.	03	23	.	38	2	.	60	154
Farmer	60	.	47	32	.	56	6	.	98	43
NA	43	.	10	51	.	72	5	.	17	58

Obviously, voters from farmer and blue collar worker households were especially supportive of President Truman, while voters of upper white collar background largely supported the Republican Candidate Dewey.

gt2 <- genTable(percent(vote3)~relig3,data=vote.48)
ftable(gt2,row.vars=2)

relig3	Truman			Dewey			Other			N
Protestant	44	.	71	50	.	98	4	.	31	255
Catholic	66	.	02	33	.	98	0	.	00	103
Other,none	68	.	18	29	.	55	2	.	27	44
NA	NaN			NaN			NaN			0

This table shows that Catholics and adherents of other denominations were more supportive of Truman than of Dewey.

gt3 <- genTable(percent(vote3)~race2,data=vote.48)
ftable(gt3,row.vars=2)

race2	Truman			Dewey			Other			N
White	51	.	33	45	.	48	3	.	19	376
Black	64	.	71	35	.	29	0	.	00	17
NA	88	.	89	11	.	11	0	.	00	9

African Americans apparently supported Truman by a large majority. The number of members of this group in the sample is very small, however, so that such an inference would be very shaky.

gt4 <- genTable(percent(vote3)~total.income,data=vote.48)
ftable(gt4,row.vars=2)

total.income	Truman			Dewey			Other			N
UNDER $500	50	.	00	50	.	00	0	.	00	8
$500-$999	61	.	54	38	.	46	0	.	00	13
$1000-1999	64	.	41	32	.	20	3	.	39	59
$2000-2999	66	.	99	30	.	10	2	.	91	103
$3000-3999	47	.	52	48	.	51	3	.	96	101
$4000-4999	45	.	83	50	.	00	4	.	17	48
$5000 AND OVER	31	.	82	68	.	18	0	.	00	66
NA	50	.	00	25	.	00	25	.	00	4

The table of percentage of vote by income suggests that income had some considerable influence on the choice either of Truman or of Dewey, but the unequal distribution of income categories warrants a more refined analysis that takes into account the uncertainty about the vote percentages. Therefore, the percentages of support for Truman broken down by income shown with confidence intervals:

## For knitr-ing, we use ```{r, ftable.digits=c(2,2,2)} here.
inc.tab <- genTable(percent(vote3,ci=TRUE)~total.income,data=vote.48)
ftable(inc.tab,row.vars=c(3,2))

total.income	vote3	Percentage			lower			upper
UNDER $500	Truman	50	.	00	15	.	70	84	.	30
	Dewey	50	.	00	15	.	70	84	.	30
	Other	0	.	00	0	.	00	36	.	94
$500-$999	Truman	61	.	54	31	.	58	86	.	14
	Dewey	38	.	46	13	.	86	68	.	42
	Other	0	.	00	0	.	00	24	.	71
$1000-1999	Truman	64	.	41	50	.	87	76	.	45
	Dewey	32	.	20	20	.	62	45	.	64
	Other	3	.	39	0	.	41	11	.	71
$2000-2999	Truman	66	.	99	57	.	03	75	.	94
	Dewey	30	.	10	21	.	45	39	.	92
	Other	2	.	91	0	.	60	8	.	28
$3000-3999	Truman	47	.	52	37	.	49	57	.	70
	Dewey	48	.	51	38	.	45	58	.	67
	Other	3	.	96	1	.	09	9	.	83
$4000-4999	Truman	45	.	83	31	.	37	60	.	83
	Dewey	50	.	00	35	.	23	64	.	77
	Other	4	.	17	0	.	51	14	.	25
$5000 AND OVER	Truman	31	.	82	20	.	89	44	.	44
	Dewey	68	.	18	55	.	56	79	.	11
	Other	0	.	00	0	.	00	5	.	44
NA	Truman	50	.	00	6	.	76	93	.	24
	Dewey	25	.	00	0	.	63	80	.	59
	Other	25	.	00	0	.	63	80	.	59

Occupational class is more evenly distributed in the sample, thus it may be possible to obtain more precise estimates of the percentages of support for Truman for occupational classes:

occup.tab <- genTable(percent(vote3,ci=TRUE)~occup4,data=vote.48)
ftable(occup.tab,row.vars=c(3,2))

occup4	vote3	Percentage			lower			upper
Upper white collar	Truman	19	.	77	11	.	96	29	.	75
	Dewey	77	.	91	67	.	67	86	.	14
	Other	2	.	33	0	.	28	8	.	15
Other white collar	Truman	49	.	18	36	.	14	62	.	30
	Dewey	50	.	82	37	.	70	63	.	86
	Other	0	.	00	0	.	00	5	.	87
Blue collar	Truman	74	.	03	66	.	35	80	.	75
	Dewey	23	.	38	16	.	94	30	.	86
	Other	2	.	60	0	.	71	6	.	52
Farmer	Truman	60	.	47	44	.	41	75	.	02
	Dewey	32	.	56	19	.	08	48	.	54
	Other	6	.	98	1	.	46	19	.	06
NA	Truman	43	.	10	30	.	16	56	.	77
	Dewey	51	.	72	38	.	22	65	.	05
	Other	5	.	17	1	.	08	14	.	38

The upper and lower white-collar and blue-collar classes are quite distinct with regard to the percentages of support for Truman. The point estimates of the percentages are outside the confidence intervals of the respective other occupational classes, the confidence intervals do not even overlap. However, it is not clear whether farmers are distinct from the blue-collar and lower white-collar classes.

Logit modelling of candidate choice

In the following we conduct a logit analysis of the vote for Truman. First, we assign non-standard contrasts the categorical predictors. Here, the function contr is used to assign treatment (dummy) contrasts to occup4 and total.income with baseline category 3 and 4, respectively.

vote.48 <- within(vote.48,{
  contrasts(occup4) <- contr("treatment",base = 3)
  contrasts(total.income) <- contr("treatment",base = 4)
  })

We now fit some logistic regression models of the impact occupational class, income, and religious denomination on the vote choice supporting Truman. The contrasts of the occupational class and income factors are such that they compare the choices of the members of the blue-collar class with all other classes and the middle income group ($ 2000-2999) with the other income groups. The religious denomination factor compares Protestants with Catholics and those with other or no denominations.

model1 <- glm((vote3=="Truman")~occup4,data=vote.48,
              family="binomial")
model2 <- glm((vote3=="Truman")~total.income,data=vote.48,
              family="binomial")
model3 <- glm((vote3=="Truman")~occup4+total.income,data=vote.48,
              family="binomial")
model4 <- glm((vote3=="Truman")~relig3,data=vote.48,
              family="binomial")
model5 <- glm((vote3=="Truman")~occup4+relig3,data=vote.48,
              family="binomial")

First, we use mtable to construct a comparative table of the estimates of model1, model2, and model3. We thus can compare the impact of occupational class and income on the choice of candidate Truman.

mtable(model1,model2,model3,summary.stats=c("Nagelkerke R-sq.","Deviance","AIC","N"))


Calls:
model1: glm(formula = (vote3 == "Truman") ~ occup4, family = "binomial", 
    data = vote.48)
model2: glm(formula = (vote3 == "Truman") ~ total.income, family = "binomial", 
    data = vote.48)
model3: glm(formula = (vote3 == "Truman") ~ occup4 + total.income, family = "binomial", 
    data = vote.48)

===============================================================================
                                             model1      model2      model3    
-------------------------------------------------------------------------------
  (Intercept)                                1.047***    0.708***    1.316***  
                                            (0.184)     (0.210)     (0.268)    
  occup4: Upper white collar/Blue collar    -2.448***               -2.328***  
                                            (0.327)                 (0.357)    
  occup4: Other white collar/Blue collar    -1.080***               -1.015**   
                                            (0.315)                 (0.323)    
  occup4: Farmer/Blue collar                -0.622                  -0.792*    
                                            (0.362)                 (0.383)    
  total.income: UNDER $500/$2000-2999                   -0.708      -0.662     
                                                        (0.737)     (1.056)    
  total.income: $500-$999/$2000-2999                    -0.238       0.912     
                                                        (0.607)     (1.143)    
  total.income: $1000-1999/$2000-2999                   -0.115       0.144     
                                                        (0.343)     (0.440)    
  total.income: $3000-3999/$2000-2999                   -0.807**    -0.527     
                                                        (0.289)     (0.338)    
  total.income: $4000-4999/$2000-2999                   -0.875*     -0.509     
                                                        (0.358)     (0.411)    
  total.income: $5000 AND OVER/$2000-2999               -1.470***   -0.535     
                                                        (0.337)     (0.405)    
-------------------------------------------------------------------------------
  Nagelkerke R-sq.                           0.246       0.085       0.274     
  Deviance                                 404.190     524.433     390.551     
  AIC                                      412.190     538.433     410.551     
  N                                        344         398         340         
===============================================================================
  Significance: *** = p < 0.001; ** = p < 0.01; * = p < 0.05

mtable returns an object of class "mtable". When formatted it looks close to the requirements of typical social science publications. Yet at least we want to change the technical variable names into non-technical ones, for which we can use relabel:

relabel(mtable(
            "Model 1"=model1,
            "Model 2"=model2,
            "Model 3"=model3,
            summary.stats=c("Nagelkerke R-sq.","Deviance","AIC","N")),
          UNDER="under",
          "AND OVER"="and over",
          occup4="Occup. class",
          total.income="Income",
          gsub=TRUE
          )


Calls:
Model 1: glm(formula = (vote3 == "Truman") ~ occup4, family = "binomial", 
    data = vote.48)
Model 2: glm(formula = (vote3 == "Truman") ~ total.income, family = "binomial", 
    data = vote.48)
Model 3: glm(formula = (vote3 == "Truman") ~ occup4 + total.income, family = "binomial", 
    data = vote.48)

====================================================================================
                                                 Model 1     Model 2     Model 3    
------------------------------------------------------------------------------------
  (Intercept)                                     1.047***    0.708***    1.316***  
                                                 (0.184)     (0.210)     (0.268)    
  Occup. class: Upper white collar/Blue collar   -2.448***               -2.328***  
                                                 (0.327)                 (0.357)    
  Occup. class: Other white collar/Blue collar   -1.080***               -1.015**   
                                                 (0.315)                 (0.323)    
  Occup. class: Farmer/Blue collar               -0.622                  -0.792*    
                                                 (0.362)                 (0.383)    
  Income: under $500/$2000-2999                              -0.708      -0.662     
                                                             (0.737)     (1.056)    
  Income: $500-$999/$2000-2999                               -0.238       0.912     
                                                             (0.607)     (1.143)    
  Income: $1000-1999/$2000-2999                              -0.115       0.144     
                                                             (0.343)     (0.440)    
  Income: $3000-3999/$2000-2999                              -0.807**    -0.527     
                                                             (0.289)     (0.338)    
  Income: $4000-4999/$2000-2999                              -0.875*     -0.509     
                                                             (0.358)     (0.411)    
  Income: $5000 and over/$2000-2999                          -1.470***   -0.535     
                                                             (0.337)     (0.405)    
------------------------------------------------------------------------------------
  Nagelkerke R-sq.                                0.246       0.085       0.274     
  Deviance                                      404.190     524.433     390.551     
  AIC                                           412.190     538.433     410.551     
  N                                             344         398         340         
====================================================================================
  Significance: *** = p < 0.001; ** = p < 0.01; * = p < 0.05

The comparison of the pseudo-R-Square values of model 1 and 2 suggests that occupational class has a stronger influence on a preference for Truman than household income. Indeed, if occupational class is taken into account, the effect of income is no longer statistically significant as the column corresponding to model 3 indicates.

Second, we compare the effect of occupational class and religious denomination on the preference for Truman along the same lines as above. We use mtable to collect the estimates of model1, model4, and model5 into a common table.

relabel(mtable(
              "Model 1"=model1,
              "Model 4"=model4,
              "Model 5"=model5,
              summary.stats=c("Nagelkerke R-sq.","Deviance","AIC","N")),
            occup4="Occup. class",
            relig3="Religion",
            gsub=TRUE
            )


Calls:
Model 1: glm(formula = (vote3 == "Truman") ~ occup4, family = "binomial", 
    data = vote.48)
Model 4: glm(formula = (vote3 == "Truman") ~ relig3, family = "binomial", 
    data = vote.48)
Model 5: glm(formula = (vote3 == "Truman") ~ occup4 + relig3, family = "binomial", 
    data = vote.48)

====================================================================================
                                                 Model 1     Model 4     Model 5    
------------------------------------------------------------------------------------
  (Intercept)                                     1.047***   -0.213       0.698**   
                                                 (0.184)     (0.126)     (0.216)    
  Occup. class: Upper white collar/Blue collar   -2.448***               -2.385***  
                                                 (0.327)                 (0.337)    
  Occup. class: Other white collar/Blue collar   -1.080***               -1.098***  
                                                 (0.315)                 (0.326)    
  Occup. class: Farmer/Blue collar               -0.622                  -0.346     
                                                 (0.362)                 (0.374)    
  Religion: Catholic/Protestant                               0.877***    0.685*    
                                                             (0.243)     (0.292)    
  Religion: Other,none/Protestant                             0.975**     1.191**   
                                                             (0.347)     (0.441)    
------------------------------------------------------------------------------------
  Nagelkerke R-sq.                                0.246       0.060       0.281     
  Deviance                                      404.190     537.711     393.105     
  AIC                                           412.190     543.711     405.105     
  N                                             344         402         344         
====================================================================================
  Significance: *** = p < 0.001; ** = p < 0.01; * = p < 0.05

A comparison of the pseudo-R-squared values suggests that also the effect of religious denomination is weaker than that of occupational class. However, as the third column in the above table indicates the effect of religious denomination remains statistically significant.

National Election Studies, 1948: Post-Election Study [dataset]. Ann Arbor, MI: University of Michigan, Center for Political Studies [producer and distributor], 1999. ANES Dataset ID: 1948.T; ICPSR Study Number: 7218. These materials are based on work supported by the National Science Foundation under Grant Nos.: SBR-9707741, SBR-9317631, SES-9209410, SES-9009379, SES-8808361, SES-8341310, SES-8207580, and SOC77-08885. Any opinions, findings and conclusions or recommendations expressed in these materials are those of the author(s) and do not necessarily reflect those of the National Science Foundation.↩