Iowa State University

Janet Nieuwsma Melby

As a scientist, Janet Nieuwsma Melby's research focuses on adolescent substance use, adolescent academic performance and observational methodology. Data for this research are mainly from the Iowa Youth and Families Project, the Family Transitions Project, the Family and Community Health Project and the Drug Abuse in Rural America Project.

Melby and colleagues found adolescent males with tobacco-using peers and/or a sibling who used tobacco more likely to use tobacco concurrently and in the future. Harsh/inconsistent parenting was associated with higher adolescent males' tobacco use, whereas nurturant/involved parenting was associated with lower adolescent tobacco use. Most associations operated directly and all had indirect effects through peers.

Four patterns of substance use from 8th through 12th grades were identified in a  series of analyses by Melby and Vargas.  Abstainers showed no or low smoking or alcohol use; for this group, family factors were important and school performance higher. Increasers accelerated use over time with a sudden higher rate of consumption from 10th to 12th grades. Heavy users initiated early, had a high level of use, and increased use over time. Both increasers and heavy users exhibited lower school performance and parents who smoked and had alcohol problems. The fourth group, decreasers, showed heavy use during 8th through 10th grades and use declined by 12th grade.  Decreasers and heavy users showed higher substance use among siblings and peers. In general, low tobacco and alcohol use were related to low or no parental or sibling use. Male gender, higher deviant values, lower school performance and friends and/or siblings who used tobacco characterized youth most at risk for tobacco involvement.

Important regional, gender and race differences in youth substance use were identified among participants from Iowa and the Lower Mississippi Delta in research conducted by Melby and Dan Hoyt. Results suggest the need to consider these differences in intervention efforts.

In the area of academic performance, Melby's longitudinal analyses revealed that parenting practices actually foster changes in academic accomplishment across time. Setting and positively reinforcing appropriate behavioral  standards (parent management) increased adolescents' academic performance, whereas hostility (negative emotional affect) decreased academic performance. Parental hostility also increased adolescents' problems at school, operating through their perception of being mistreated, thus supporting a social-contextual perspective.

Observational methodology is a big part of Melby's work at the institute. The Iowa Family Interaction Rating Scales, that she and others produced, provides a useful set of measures for studying social processes and individual development. With this system, trained observers supply valuable information on family processes and relationships, especially in areas that may be difficult for participants to assess objectively. In conjunction with self- and family- member-reports, observer ratings enable researchers to employ multi-informant techniques in testing hypotheses, thus reducing method-variance problems. Intraclass correlations (ICCs) for observer ratings range from .55 to .85 for the single scales used in most analyses. In general, ICCs are higher for composite measures than for single ratings. Structural equation modeling techniques also provide a useful framework for evaluating quality of observer and family-member assessments within a single analytic context.

Analyses using observational measures of martial hostility and relationship quality demonstrated acceptable internal consistency of items, substantial reliability for constructs, and support for the validity of observational measures. Assessments of marital warmth/prosocial behaviors supported the reliability of global ratings and illustrated that evaluations of validity must consider task context (discussion or problem solving). When observer ratings of adolescent behaviors were validated against self-, sibling-, and parent-reports of behaviors and evaluations of the relationship, age of sibling (younger or older than adolescent) and type of task context (parents present or absent) were important considerations. Acceptable inter-observer reliability, internal consistency and test/retest reliabilities have been obtained for composite  measures of parental warmth, hostility and discipline. Preliminary results suggest the absence of systematic bias in ratings of Caucasian and African-American parent-child dyads by African-American and Caucasian observers, although some mean differences in ratings of hostility and warmth/support were observed.

As the Director of the Observation Unit as ISBR, Janet Nieuwsma Melby oversees approximately 30 full and part-time staff members who specialize in behavioral assessments of children, adolescents, young adults, and parents engaged in temperament, activity, or discussion tasks.  More information about the work of the Observation Unit is available through the unit link on this website.

Ames, Iowa 50011, (515) 294-4518. Published by: ISBR Computer Support, webadmin@isbr.iastate.edu. Copyright © 1998-2004, Iowa State University. All rights reserved.