The Millennials Friendship Study

Purpose: The current study has two main objectives: (1) to develop and evaluate new methodological techniques for harmonizing measures for substance use and related disorders across a broad range of measurement contexts and (2) to evaluate a series of hypotheses concerning the social and environmental factors that lead young-adult college students towards or away from substance use behavior. However, in accomplishing these objectives, we also included measures that have to do with friendships (history and quality) and well-being (the Ryff scales) as well as social media use, gratitude, and health behaviors (sleep, substance use, depression). The Ryff scales also consider: autonomy, environmental mastery, personal growth, positive relationships with others, purpose in life, and self-acceptance. The relationship quality scale (the NRI) assess friendship quality in terms of affection (liking and loving), companionship, conflict, instrumental aid, antagonism, intimate disclosure, nurturance, and reassurance of worth. Sample: The study sample was comprised of 923 pairs college student friend dyads (total =1846) from a single university. Inclusion criteria for target participants were being between 18-26 years of age, currently enrolled as a university student, and reports of any alcohol use in the past year at the point of screening. Inclusion criteria for peer participants included being over age 18 and never having participated in the study before. Male and ethnic minority students were oversampled. Students were recruited through two sources. First, we sent invitations to 8,199 undergraduate students whose contact information we obtained through the registrar’s office or who participated in the a previous study. Second, we had an additional 116 individuals contact us asking to participate in the study after hearing about it from a friend. Out of the 1,468 who completed our pre-screen from either source, 1,270 were eligible to participate; ineligibility was due to age (1), non-student status (1) no alcohol use in the past year (17) and previous involvement in the study (3). Of those eligible, 949 scheduled sessions for the study. A total of 923 students completed the first session and 901 dyads (or 922 targets and 902 peers) completed both visits. This study is a companion to the REAL-U study, that used a similar battery with college students but did not use a paired friendship design. (But could provide replication data and for a smaller subsample longitudinal follow-up data). Together the two studies include 2,600 participates, with 1584 European American, 660 Black/African Americans, 240 Asian/Asian Americans, and 266 Latinx students. Procedures. For the MFS study, selected students were first sent a personalized email containing a brief description of the study, a link to the study’s website, and a link to a Qualtrics-based prescreen survey. The prescreen evaluated the target’s eligibility and, for qualifying individuals, allowed them to nominate up to four friends to participant in the study with them as their peer. The first-choice friend, (the status of whom was designated by the order in which the target nominated their friends), automatically received a pre-screen survey of their own, which evaluated their eligibility. If the peer was determined to be eligible, their completed survey triggered an automatic email from Qualtrics that provided both the target and the peer with sign up instructions for the study’s research portal. The pair was instructed to sign up for two testing sessions, exactly two weeks apart that they were both available to attend together. At the first testing session, participants completed consent procedures and a computerized battery of surveys. Participants were randomly assigned to one of five conditions, which determined the batteries to be completed at session one and session two. There were four batteries: Battery A and Battery B which were designed for target participants and Battery C and Battery D, designed for their peer counterpart. In the first condition, target participants would complete Battery A both sessions, and their peer would complete Battery C at both sessions. In the second condition, targets would complete Battery B at both sessions and their peers would complete Battery D at both sessions. In the third condition, the target would complete Battery A at visit one, then Battery B at visit two, and their peers would complete Battery C at visit one, then Battery D at visit two. In the fourth condition, targets would complete Battery A at both sessions while their peers would complete Battery D at both sessions. In condition five, target participants would complete Battery B at visit one, then Battery A at visit two, while their peers would complete Battery C at visit one and Battery D at visit two. In general, target batteries (A and B) primarily consisted of questions about the target participant him- or herself while the peer batteries (C and D) asked questions about the peers as well as about the target friend who participated with them. Procedures for sessions one and two took approximately 60 minutes to complete. At the end of the second testing session, participants were invited to participate in a second consent procedure permitting the study team to access their student adjudication data as protected by FERPA. Participants received a $20 incentive for completion of the first testing session and a $25 incentive for completing the second testing session (regardless of whether they provided consent for their adjudication data). If participants consented to waiving their FERPA rights, they were entered into one of four lottery drawings for $100 each. Procedures for harmonizing and parsing data are the primary focus of the data design but can be leveraged for a variety of questions. UNC-CH IRB approval #15-2710.