I have chosen two variables I think are causally related: educational attainment and attitudes toward pornography laws. I believe that people who are better educated will be more tolerant of those who want to buy pornographic material than those with less education. I know from reading and my life experiences that better educated people are more likely to support civil liberties guarantees, and the right to purchase pornographic material is considered one of those civil liberties. Less educated people, on the other hand, may not have been exposed to as great a diversity of different people (and hence different viewpoints). As a consequence, they may be less likely to agree that their fellow citizens should be able to purchase pornographic material.

My theory is that education is related to tolerance of pornography. My formal hypothesis is: the more educated an individual, the more tolerant he/she will be toward the distribution of pornography.

To conceptualize education, I am talking about the number of years an individual remains in school. My hypothesis states that the longer an individual stays in school, the more tolerant he/she will be toward pornography. Therefore, in my measure of education I need a variable that tells me how long an individual remained in school. In particular, I expect that those who have a college degree (16 or more years of schooling) will be more tolerant than those with at least some college (13-15 years), who will in turn be more tolerant than those with a high school degree (12 years), who will be more tolerant than those with less than a high school degree (0-11 years of schooling). Thus, in conceptualizing education I am talking about years of school completed.

With respect to tolerance of pornography, I want to measure degree of tolerance, not just whether a respondent is for or against the availability of pornography. It may be that people tolerate pornography if adults buy it, but not when it is made available to children. Thus I want a measure of tolerance that distinguishes between pornography available to adults and that available to children. I decided to ask people their views on pornography laws, distinguishing between those who would be less likely to support laws against pornography, and those who would be more likely to support pornography laws.

To operationalize these concepts, we specify the actual indicators we will use to measure our concepts. The question used to operationalize education was "educ": "What is the highest grade in elementary school, high school, or college that you finished and got credit for?" Possible responses range from 00 (no formal schooling) to 20 (8 or more years of college) [note: this question comes directly from the General Social Survey-see the "educ" listing on the GSS website]. This operationalization means that the respondent's schooling ranges from a score of 0 to a score of 20, with higher numbers indicating greater achieved education. I recoded this to an ordinal categorical variable: "1"=less than high school degree (0-11 yrs), "2"=high school degree (12 years); "3"=some college (13-15 yrs), "4"=college grad or more (16+ yrs).

The question used to operationalize tolerance of pornography was "pornlaw": "Which of these statements comes closest to your feelings about pornography laws?" Possible responses are "1" (there should be laws against the distribution of pornography whatever the age); "2" (there should be laws against the distribution of pornography to persons under 18); "3" (there should be no laws forbidding the distribution of pornography). Note that a code of "1" would indicate least tolerance of pornography; "2" would indicate moderate tolerance of pornography, in the sense that adults should be able to have access to pornography if they want it, but children should not; "3" indicates greatest tolerance of pornography--there should be no laws restricting pornography.

Given this operationalization, my expectation is that the more education one has, the more tolerant he or she should be. I ran this analysis using the 1998 GSS and found the following results:

Table 1. Tolerance of Pornography by Years of School Completed, U.S. 1998
(in percentages)

[Note: the table is percentaged in the direction of causality: education "causes" pornography attitudes. "Percentage down, compare across."]

Analysis: What do these results tell us about the relationship between education (the independent variable) and tolerance of pornography (the dependent variable)? These data support my hypothesis that those with more education are more tolerant of pornography. This is reflected in the fact that they are less likely to support laws against pornography for everyone regardless of age. While 45 percent of those with less than a high school degree favor total restriction of pornography for everyone, 32 percent of those with at least a college degree favor total restriction. The effect of education is just the opposite when pornography laws protecting children are considered: 65 percent of the most educated support laws against pornography for persons less than 18 compared with 49 percent of those with less than a high school degree. Few Americans are fully tolerant of pornography, however: regardless of educational attainment, less than six percent of the U.S. population in 1998 believed that there should be no laws restricting pornography.

Reformulation of theory: The results of my analysis support my theory that education is positively related to tolerance of pornography. Less educated people were more likely to support laws against pornography for everyone. More educated people supported laws designed to protect children, but were more tolerant on issues of choice for adults. In working through this analysis, however, it occurred to me that tolerance of pornography may vary not only by education, but also by sex, or perhaps religion. It may be that women are less tolerant of pornography than men because they view the very existence of the pornography industry as exploitative, since most pornography is directed toward a male audience. On the other hand, people of different religions may vary in their tolerance of pornography. This thought process leads me to want to rerun the analysis described above, adding the additional control variable(s) of sex and/or religion of respondent.

And the process goes on and on and on . . .