Utilizing Technology for Longitudinal Data Collection With Homeless Youth

Background: Because homeless youth are often transient and lack stable contact information, little longitudinal research has been conducted with this group. Accordingly, researchers’ abilities to draw causal conclusions regarding factors that predict youth outcomes are limited. While technology may improve study retention, further research is needed to identify the best methods for maintaining longitudinal contact with homeless youth. The current study investigated the utility of various technological approaches. Research questions addressed, among homeless youth: 1) which contact methods are associated with successful longitudinal data collection, 2) how many contact attempts are necessary, and 3) what youth characteristics are associated with study retention?

Methods: As part of a clinical trial for homeless youth, data were collected from youth ages 18-24 (N=66) accessing services in a homeless youth shelter. Participants were given active cell phones (pre-paid for 3 months), and participants’ e-mail and Facebook information were collected. Youth were contacted via phone, text, e-mail, or Facebook at four time points: pretest, 1-week posttest, 6-week follow-up, and 3-month follow-up. Interviewers documented each contact attempt, and analyses described number of attempts to reach youth, technology methods used, and whether youth were retained in the study at each follow-up. Qualitative field notes supplemented quantitative data by describing challenges encountered in maintaining contact with youth using various technologies.

Results: Approximately 96% of youth were retained at 6-weeks and 94% at 3-months. On average, it took 3.65 (SD=2.3) back-and-forth exchanges between interviewers and youth over 2.18 (SD=2.7) days to successfully engage youth in 6-week follow-up interviews. Greater time was required at 3-month follow-up, averaging 3.39 (SD=3.4) exchanges over 3.45 (SD=3.4) days. Although a combination of contact methods was used, cell phone contact (calls or texts) was most successful, with e-mail and Facebook messaging useful only when phones were lost or stolen. Youth preferred participating in 6-week interviews in-person (55%) versus by phone (45%). However, by 3-month interviews, youth preferred phone (61%) to in-person interviews (39%). Intervention and control groups were retained in the study at similar rates. Bivariate analyses revealed youth who dropped out of the study were less likely to use other shelter services (e.g., GED, job training) and had been homeless nearly twice as long as youth who were retained.

Conclusions: Longitudinal research with homeless youth is feasible if resources are available for cell phones and service, and persistent exchanges are employed to achieve follow-up.  Cell phones demonstrate particular promise in retaining homeless youth as youth become increasingly transient and difficult to meet in-person over time. Qualitative field notes revealed that, although youth highly valued being given phones, several lost, gave away, or had them stolen, requiring back-up contact methods. E-mail and Facebook elicited inconsistent response from youth who changed e-mail addresses or forgot passwords, though they proved important contact methods if phones were lost. Facebook introduced new challenges (e.g., given the amount of personal information outside the study focus posted on youths’ walls). Future longitudinal research with this difficult-to-track population would benefit from providing phone access while also collecting back-up contact methods.