Research Paper on "Interpreting Social Data"

Research Paper 7 pages (1825 words) Sources: 0 Style: APA

[EXCERPT] . . . .

Statistical Relationships in the General Social Survey:

Religion, Class, and Culture

In general, people believe that there are correlations between religion, class, and cultural beliefs in the United States. For example, it would be uncontroversial to assert that atheists might have a more liberal social stance on abortion. Some might believe that people in higher income brackets are less likely to be extreme in their religious faith, or that they are more likely to have been divorced. People also have a variety of beliefs about the causal direction of these relationships - does a strong religious faith cause certain social beliefs, or do conservative social beliefs make it easier to fit in among members of a conservative faith community? We are also aware - particularly through the news media - of inconsistencies between people's actions and their beliefs; for example, politicians who spend years eradicating corruption and fighting for transparency in government are later arrested for soliciting prostitutes (Elliot Spitzer). We may be curious how common these moments of hypocrisy are, whether they are really newsworthy or if they are endemic to people with strongly-held beliefs.

In the following analysis, I will explore some of the relationships between cultural, religious, and class/socioeconomic variables in the General Social Survey data from 2004. The GSS provides a wide variety of data to explore, of nominal, ordinal, and interval formats. The variables of most interest to me in uncovering relationships between religion, class and culture are captured in the following survey items: Religion, Party identification, Class, Biblical literalism, Political views, Beliefs about teaching children the value of hard work, Income, and whether or not respondent had viewed an X-rated film in the last year. The first three I will treat as nominal variables, the next three as ordinal variables, and the last two as interval or ratio variables, as explained below.

Some of the hypotheses I hope to partially test by exploring this small data set relate to the social and religious content of political conservatism. Intuitively, religious and social and political conservatism may all go together in individuals. However, the data may show that politically conservative individuals are actually more likely to be religiously liberal. Judicious use of the CROSSTABS analysis can shed further light on the potential bivariate relationship between positions on both political and religious scales of liberality. One specific hypothesis that can be answered by these data is: are individuals who are more politically conservative (e.g. identify as having conservative political views) also more likely, if Christian, to believe that the Bible is the literal word of God? Moreover, are these the same people who believe in teaching children the value of hard work - presumably also a conservative cultural value?

First, however, I will explore some general trends in the data that may inform the intuitions that I expressed above. Using the FREQUENCIES analysis procedure, we see an interesting distribution of responses to the question, "How often do you attend religious services?" In Figure 1 (See appendix of tables and figures at the end of this document). We can see in this histogram that although SPSS has generated a unimodal normal curve that vaguely fits the data, the rough peaks at the beginning and end of the histogram indicate that the data might be better described by a bimodal distribution. In other words, there are quite a few people who attend religious services frequently, and there are quite a few people who attend them irregularly or not at all, but there are fewer people in the middle categories than at the ends of the spectrum.

Another interesting, and perhaps troubling, finding using the FREQUENCIES procedure is the number of data points missing from many of the more sensitive questions. For example, if respondents were asked whether or not they viewed an x-rated movie in the last year, less than a third have responded to this question. However, similar proportions exist for the question about the importance of teaching children the value of hard work, which is presumably not an offensive or controversial thing to admit. Using the CROSSTABS analysis may help to illuminate which groups of respondents were matched with which questions, in order to reduce spurious results. For example, if these data include respondents to two non-overlapping groups of questions, a bivariate analysis will produce very low correlations between variables that may, in fact, be highly related in the population in question. Other questions (marital status, party affiliation, social class, age) seem to be universally asked and would produce reliable results in a bivariate analysis. Proportions of missing data for each question, ranked roughly by "controversiality," appear in Table 1.

In terms of answering questions about the hypothesis that religious, social, and political conservatisms go hand in hand, the frequencies of answers to the relevant questions have little to say. Since frequency counts address only a single variable at a time, it is impossible to get an idea of bivariate patterns, but one can examine general trends in the population. Some general trends that emerge in the variables I have chosen to examine are reflected in the histograms represented in Figures 2 through 5. Figure 2 shows a histogram of whether respondents believed the Bible is the literal word of God, the non-literal but inspired word of God, the work of entirely human forces, or belonged to a category not mentioned. In general, the sample population seems to show a strong trend towards believing that the writings in the Bible were at least somewhat influenced by God - if not directly given, then "inspired." Only 16.3% of the sample were of the opinion that the Bible is an ancient book written by humans with no particular divine inspiration. Figure 3 depicts the frequencies of party identification - from Strong Republican through Independent to Weak, Moderate, and Strong Democrat. It is possible that these categories should be treated more nominally, but since the two-party system allows for a certain natural polarity in the data, I will treat these values as ordinal, especially since they are phrased in terms of strength of alignment. In this sample, all participants responded to this question. It is therefore unsurprising that we see a balanced proportion of Democrats, Republicans, and Independents. The existence of these parties (and the third "independent" designation) probably pulls respondents to think of themselves as belonging to one of the categories instead of hanging out in the nebulous space between affiliations. This explains the three-peaked structure of this question's frequency distribution. Figure 4 shows the distribution of responses to a question of how important the value of hard work should be in forming a child's ethics. Clearly, very few people believe that the value of hard work should be the first thing on a parent's mind, but there are plenty of respondents who believe it should be second or third on the list of important values your child should learn - about 20% across those two categories. If valuing hard work is perceived to be a conservative trait, then this distribution is in accord with the next frequency distribution, in Figure 5. Figure 5 depicts the frequency of respondents labeling their own political views from Extremely Liberal to Extremely Conservative. We see the strong bias towards regression to the mean very clearly with this question, in the high middle peak representing the frequency of the response "Moderate, middle of the road, neither liberal nor conservative." People who associate moderation with reasonable behavior will be drawn towards describing their political stance as more moderate than it may actually be; also, people who do not think a great deal about politics may be inclined to call themselves moderate instead of claiming political ignorance. However, when we see past the proportion of respondents who regress to the mean, we can see that there is an underlying skew to the data in the direction of more conservative beliefs. People are claiming the label of "very conservative" more than they are claiming any of the labels "slightly liberal," "liberal," or even "slightly conservative." Any firmer correlations between conservatism in the population and specific beliefs about values will be discussed below in the CROSSTABS analyses.

Several of the variables I have examined are good candidates for bivariate relationships with each other. I will discuss two in particular: the relationship between degrees of poverty and belief that hard work is an important value for children to learn; the relationship between reported frequency of religious service attendance and reported frequency of x-rated movie viewing. Figure 6 represents the cross-tabulation of income level (below $25,000/year) with the degree of belief that it is important for children to learn to value hard work. I have highlighted the three income categories with enough data. The trend appears to be that in lower income ranges (what many Americans might consider "extreme poverty") as well as in people living above the poverty level, individuals may be less frequent in their belief that hard work should be the number one value. Around the… READ MORE

How to Reference "Interpreting Social Data" Research Paper in a Bibliography

Related Research Papers:

Social Work Informatics USES, Merits and Demerits Research Paper

Social Work Informatics

USES, MERITS and DEMERITS

Social Work Informatics and Its Use in Child Protection

Social informatics refers to a large, varied and expanding research, which looks into the… read more

Research Paper 6 pages (1824 words) Sources: 5 Topic: Computers / IT / Internet

---

Social Psychology Cognitive Dissonance Term Paper

Social Psychology

Cognitive dissonance

This is the feeling that one has within them when they hold two conflicting feelings or thoughts within them at the same time. This usually highlights… read more

Term Paper 4 pages (1659 words) Sources: 4 Topic: Psychology / Behavior / Psychiatry

---

Criminal Justice and Criminology Interpreting Simple Data Research Proposal

Criminal Justice and Criminology

Interpreting simple data

The data collection exercise involved posting a picture of a bear on Facebook. The caption of the picture asked viewers to provide their… read more

Research Proposal 7 pages (2030 words) Sources: 6 Topic: Mathematics / Statistics

---

Social Interactions Between Alternative Therapists and Patients Thesis

Social Interactions Between Alternative Therapists and Patients

The goal of the research in this work has as its focus interactions that take place among natural and social groups. This work… read more

Thesis 14 pages (4392 words) Sources: 5 Style: APA Topic: Medicine / Pharmacy

---

Social Control Theory Term Paper

Social Control Theory

All control theories play on the theme that deviance is mainly a function of the kinds of constraints to which people are exposed. The most well-known specific… read more

Term Paper 12 pages (3849 words) Sources: 1+ Topic: Sociology / Society

---

View 100+ more related papers  >>