Story ID
Part 1: Demographics
General intuition might suggest that a wide variety of demographic factors may influence the moral stories generated by our participants. One might expect that
participants growing up in different cultures, religions or socioeconomic classes have different formative experiences. Gender, IQ and education seem likely, at least
intuitively, to play a role in the moral choices that participants make. However, after collecting nearly 800 stories from 100 participants, we find just the opposite. In general, people from all backgrounds and walks of life seem to have experienced similar moral dilemmas throughout their lives and recall similar moral memories.
There are two key questions about the demographics of our participant sample.
First, are they a representative sampling of the population? To begin assessing the representativeness of our sample, we must begin by defining the population they
describe. We chose to compare them to the population from which they were drawn: the general population of southern California. In the following section, we provide a detailed characterization of the participant group. Briefly, participants were specifically matched to California state averages on ethnicity and gender. They fall within the normal range for IQ; they are both right and left handed; and their religions, secondary languages, socio- economic statuses and political beliefs are diverse. This characterization of the
participants provides the basis for our second question: How might individual differences along demographic or neuropsychological dimensions influence the moral memories that participants generated? For instance, do men and women generate different kinds of moral memories? Does age or ethnicity impact these memories? These types of exploratory questions are used to probe this very large database and the interesting findings are treated in detail in the subsequent sections.
Gender, Age and Handedness:
Our participants were diverse and representative of the general population on all of the demographic measures that we collected. The gender ratio of the participants in the
database was quite close to the state average (53% female, 47% male and 50.2% female, 49.8% male, respectively). We did not have any transgendered participants. All of our participants were between the ages of 40 and 60 years old. Participants were normally distributed (mean = 48.9 ± 5.9 years) within this range (see Figure 8). 83% of our participants were right-handed. The distribution of the others can be seen in Figure 9.
Although the proportion (and definition) of handedness in the general population is disputed184, most sources suggest that approximately 90% of people are right handed.
Education:
Participants’ education (see Figure 10) ranged from only elementary education (1 participant) to advanced graduate degrees, including medical and law school. Most participants (90%) had attended at least some college, which does make their general education level slightly higher than the local population. 56.7% of the general Californian population has attended “some college or more”185. The elevated educational status may be accounted for by the techniques used to recruit participants. All participants were recruited through computer advertising (or word of mouth) and nearly all of our participants were comfortable using computers. We do not think this difference in education level is significant to the recall of autobiographical memories.
Income:
The median income range for these participants was $30,000 - $60,000 per year.
This is in keeping with local census findings. Per capita income for Californians was
184 Oldfield, R. C. (1971). "The assessment and analysis of handedness: The Edinburgh Inventory."
Neuropsychologia 9: 97-113.
185 Bauman, K. J. and N. L. Graf (2003). Educational Attainment: 2000, US Census Bureau.
$22,711; median household income was $47,493186. Participants were asked to report individual incomes, not household incomes, during our demographic collection; however it is difficult to ascertain how accurately they followed this instruction. A significant number of our participants were not (or no longer) married, so the majority of these data are likely to be individual incomes. The distribution of incomes in our sample was somewhat skewed toward lower income ranges (Figure 11). This is likely due, in part, to participant recruitment. Our participants were paid $15/hour for their time, without reimbursement for travel. Although we would like to believe that at least some of our participants participated because of a genuine interest in research, we suspect that many of them were interested in participating to make extra money. Given this, we expected to (and did) recruit more participants with lower incomes. Despite this, 16% of our
participants made more than $60,000, supporting the claim that this is a representative sample of wage-earners in Southern California.
Occupations and Foreign Language Skills:
Our participant pool also included nearly 50 different occupations, spanning a wide range of both white and blue collar jobs, including: teachers, financial analysts, musicians, hair dressers, a farm grower, a photographer and a substance abuse counselor.
For a complete list, see Table 3. Participants were also culturally diverse. All participants spoke English as their primary language (by adolescence), but many (34%) were raised bi- or trilingual (see Figure 12). This exact statistic is not available for the Californian
186 Income data is from Welniak, E. and K. Posey (2005). Household Income: 1999, US Census Bureau.
Our participants were tested in 2004. Data from US Census 2005 are not yet completely available at the time of this writing.
population at large, but it is in keeping with the available statistic – 39.5% of Californians speak a language other than English at home.
Ethnicity:
Choosing ethnically diverse participants was of primary importance to creating a database that was representative. The ethnicity distribution of participants in our study closely matches the California state averages (see Figure 13.) Given the high proportion of Hispanic citizens (of several races) in Southern California, we chose to match our demographics to ethnicity instead of race alone. Normative data are taken from the US Census 2000187. The participant pool slightly over-represents Black and Hispanic participants (by 4.6% and 3.7%, respectively) and slightly under-represents Caucasian participants (by 6.5%).
Marital Status and Children:
The marital statuses of our participants were varied (see Figure 14). The number of married and partnered participants in the database was smaller than the average population188 (27% versus 52.3%) and the number of divorced participants was higher than average (24% versus 9.5%). Although the proportions of the marital statuses of our participants were not particularly close to the state demographics, we feel that the sample is still a good representation of the general population. All marital status categories (single, married, partnered, divorced, separated and widowed) were represented. The variance in these data is likely attributable to the size of the sample. 53% of our
187 Data taken from "US Census Bureau American FactFinder", from http://factfinder.census.gov/. All data is from US Census 2000.
188 Kreider, R. M. and T. Simmons (2003). Marital Status: 2000, US Census Bureau.
participants had at least one child (see Figure 15). (One participant had 7 children.) This number is in line with the state averages; 56.8% of families in California have children189 and the average family size is 3.43 people.
Sexual Orientation:
Our participant pool included people of most sexual orientations. The population frequency data for sexual orientation are widely disputed in current literature. A review of the literature in The New England Journal of Medicine suggests that between 2 and 6 percent of the adult population are exclusively homosexual or bisexual190. Many sources suggest that these reported numbers may be low, due to the social stigma associated with identifying as homosexual or bisexual. Our database includes participants who reported their sexual orientation as homosexual (4%) and bisexual (3%) as well as several participants who chose not to respond (6%). (See Figure 16.) One of the findings of the well-known Kinsey study191 was that sexuality cannot always be defined by a single parameter, like “sexual orientation”. We also collected attractiveness ratings and degree of sexual desire on a Likert scale (1 – 10). These findings were in concert with the sexual orientation rating (see Figure 17). Despite the controversy about the exact frequency of homosexuality in the population, the frequency of various sexual orientations in our database seems to be in line with most findings, suggesting that our participants are representative on this demographic dimension.
189 Census data is complicated here because not all of our participants would qualify as “families” by census data. Nonetheless, these numbers are probably representative.
190 Friedman, R. C. and J. I. Downey (1994). "Homosexuality." New England Journal of Medicine 331(14):
923-930.
191Gebhard, P. H. and A. B. Johnson (1998 ). The Kinsey Data: Marginal Tabulations of 1938-1963 Interviews Conducted by the Institute for Sex Research Bloomington, IN, Indiana University Press.
Religious Views:
Participants were also religiously diverse. Since religion and morality are often entangled, we felt it was particularly important to collect data about participants’
religious beliefs and practices. Data were collected on religious upbringing, current religious practices and degree of participation in religion. Figure 18 shows a trend towards religious exploration in our participants during their adult lives. A number of participants (19%) participated in multiple religions or non-conventional spiritual
practices. To gauge the effect of religion on our participants’ lives (Figure 19), they were asked to answer a free-response question about their religious upbringing and their current religious practices. These responses (usually 2-3 sentences for each question) were scored by one reader, to give each participant a ‘degree of religious influence’
rating. 92% of our participants had some degree of religious influence in their lives, either during their childhood or through their current religious practices.
Our participants were generally well-matched to state averages in the distribution of religious beliefs (Figure 20). Overall, Catholics seem to be under-represented, but we believe that some of this discrepancy is subsumed by the ‘Other Christian’ group.
Participants were performing a free-response task when asked about religion and we believe that some of them may simply have written ‘Christian’, without specifying
‘Catholic’. While reliable data about the proportion of Californians who practice many of these religions are not always available, the Center for Religious and Civic Culture192 in
192 "USC Center for Religious and Civic Culture." from
http://www.usc.edu/schools/college/crcc/demographics/lacounty.html.
conjunction with the Glenmary Research Center193 provides us with some statistics we can use to compare our sample to the general population (Table 4). In our sampling, we did not have any participants who were Muslim, although 1% of the population in California report being Muslim. We did, however, recruit several Buddhists (3%) and it appears that there are nearly 3 times as many Buddhist temples in Los Angeles County as Muslim mosques. The numbers of Muslims mosques and Hindu temples are fairly equal and we did not manage to recruit any participants from either religion, suggesting that our sampling of the population is relatively uniform.
Political Views:
Lastly, our participants were diverse in their political views. Participants’ political value systems were assessed using a questionnaire; those methods are described in
Chapter 2. Figure 21 shows that 50% of our participants were best affiliated with liberal values, 40% were moderate and 10% espoused conservative beliefs. The average rating for our overall participant pool was slightly skewed to the left (mean = 2.4 ± 0.67 on scale of 1 (Conservative) to 3 (Liberal)). In 2004, 43.2% of voters in California were registered as Democrats and 35.7 % were registered as Republicans194. Since the questionnaire used to test political values included questions that are divisive for some Republicans (like abortion), we suspect that some of our ‘moderate’ participants are likely to be affiliated with the Republican Party.
193(2000). Religious Congregations and Membership in the United States. Nashville, TN, Glenmary Research Center. Description of this group and these statistics:
http://www.usc.edu/schools/college/crcc/demographics/
194 "Report of Registration: January 2, 2004." from http://www.ss.ca.gov/elections/ror/regstats_01-02- 04.pdf.
One caveat worth noting is that our participants were selected to have a somewhat uniform cultural context. They were all residents of Southern California who had lived in the area for at least 15 years. The intent of this restriction, originally, was to ensure that our database of memories had some uniformity in terms of social mores, societal rules and general tenets of behavior. Given the diversity demonstrated in the demographics, we feel that even given this restriction, we have constructed a broad, representative sample of Southern Californian society.