General Purpose, Questions, and Sample

Students in the fall 2004 methods class suggested that we study friendship. Social scientists have thoroughly documented the importance of friendship throughout the life course. In Friendship Processes, psychologist Beverley Fehr (1996) cites research findings indicating, for example, that people associate more pleasure with the presence of friends than with being alone or with family; the majority of adults’ social networks consists of friends; and more socially integrated individuals—as assessed by friendships—tend to live longer. Friendship is particularly important to college students, especially when they live on campus, as almost every activity in which they engage—eating, sleeping, attending class, participating in sports, attending campus events, partying, and so forth—takes place in the presence of peers. Noting the centrality of friendship among the Rutgers undergraduates he studied in the 1970s and 1980s, anthropologist Michael Moffatt (1989) stated that “friendship has been the core relationship in undergraduate culture for two centuries.” Fehr also cites one study which showed that college students more often named friends as what makes their lives meaningful than any other source of meaning, including family.

To measure the structure of students’ friendship networks, the fall survey began by asking students to identify their four closest friends and their relationships with one another. After asking about specific characteristics of each of the four friends, we then focused on activities and conversation topics with respondents’ single best friend and their closest friend who was of the opposite sex of their best friend. The survey barely scratched the surface of friendship patterns, even though interviews averaged nearly 20 minutes. Moreover, as it turned out, the data on activities and topics of conversation were of limited use, because only one-third of best friends were fellow Holy Cross students, with whom respondents regularly interact. Therefore, we conducted a second survey in the spring, which was limited to relationships with other Holy Cross students and which measured additional friendship constructs.

The two surveys asked students about the following:

• The meaning of friendship (fall, spring).
• Characteristics of friends such as gender, race, academic class, and length of relationship (fall, spring).
• Amount of time spent with best on-campus friend (spring).
• Frequency of various activities with same- and opposite-sex best friends (fall, spring).
• Frequency of various topics of conversation with same- and opposite-sex best friends (fall, spring).
• Whether respondents are romantically or intimately involved (fall, spring).
• Satisfaction with friendships (fall, spring).
• Closeness of relationship with same- and opposite-sex best friends (spring).
• Social support, social conflict, and competition with same- and opposite-sex friends (spring).
• Perception of enemies (spring).

Most of the questions were developed anew, but several were derived from instruments used in various friendship studies, in particular Ellen Berscheid, Mark Snyder, and Allen Omoto’s (1989) Relationship Closeness Inventory. Social support and social conflict items were taken from the Midlife Development in the U.S. (MIDUS) Survey. One measure of friendship closeness was Arthur Aron, Elaine Aron, and Danny Smollan’s (1992) Inclusion of Other in the Self (IOS) Scale. Three items measuring friendship satisfaction were drawn from a study by Diane Carlson Jones (1991).

The fall survey took place between October 21 and November 22; the spring survey between March 15 and April 19. As in every other HCSS, the target population and sampling frame for each survey included only students enrolled and on campus, thus excluding those who were studying away or abroad or who had taken a leave of absence. This amounted to 2,598 students in the fall and 2,569 in the spring. In addition, because the two surveys dealt with the same topic, students sampled in the fall were excluded from the sampling frame before drawing a sample in the spring. Both samples—260 in the fall and 180 in the spring—were randomly selected. A total of 229 interviews were completed in the fall and 162 interviews were completed in the spring, yielding response rates, respectively, of 88 and 90 percent. The margin of error is about 7 percent for the fall survey, 8 percent for the spring survey, and 5 percent for the two surveys combined. This means, for example, that if a reported percentage is 45 for the combined surveys, the percentage for all Holy Cross students is very likely (95 chances in 100) to fall between 40 and 50.

Of the 391 respondents in fall 2004 and spring 2005, 51.4 percent were female, 86.7 percent were white, 99.5 percent ranged in age from 18 to 22 years old, and 77.2 percent identified themselves as Catholic. Ninety-two percent of the respondents lived on campus. By academic class, 27.6 percent were first-year students, 28.9 percent second-year, 21.2 percent third-year, and 22.3 percent fourth-year.

Return to top