Meet Frank Mott

Former Principal Investigator for the NLSY79 Child/Young Adult Cohort

Frank joined the NLS team at Ohio State University in 1975. He was instrumental in launching the NLSY79 Child/Young Adult cohort, and was the long-time principal investigator for that cohort before retiring in 2010.

Due both to his longevity and his energy, Frank also holds the record for having the most entries (134) in the NLS Annotated Bibliography. In light of that fun fact, we proposed that he write an article for this website titled “Frank’s Favorite,” in which he would briefly describe his favorite NLS-based article. While we assumed that would be an easy assignment, Frank disagreed—and chose to go in a different direction.

Read Frank’s reflections on why he doesn’t want to discuss his own research, and why intergenerational links are important to him in more ways than one.

In honor of the 50th anniversary of the NLS, I was asked to choose a favorite piece of research I had produced over the years (in my case, 40 years) and discuss what I liked about it and what I learned. It would have been an especially hard task to review well over a hundred articles, pick out a few, and then to reconstruct what I had done, and evaluate them. The thought of even glancing over various and sundry articles and papers was enough to give me nightmares. And what if I have changed my mind about methods or results for earlier work that I completed? I essentially decided that it was more than I could handle.

Rather than synthesizing a few summary thoughts from work I have done, in its place I decided that it would perhaps be useful to suggest a somewhat different perspective to others, particularly those who may be searching for a new research area with a still unique data set.

The NLS has been gathering data for 50 years. It encompasses a massive amount of social, economic, demographic, and health-linked information that has covered significant segments of the lives of a broad base of the American population. While the seven cohorts in the NLS are each constrained in various ways, they have separately and cumulatively resulted in a massive amount of useful research that has contributed in endless ways to knowledge and understanding of American society.

To date, most (but certainly not all) research has focused on results from one cohort at a time. The papers I have written are typically based on data from one cohort, because that was the research cohort I was most involved with at that time, because all the others were not yet available as presently constituted, and frankly, because it was easier to do so. Of course, the more years of data available for a cohort, the greater the possibilities there are for longitudinal explorations within any single cohort.

It would be of major research utility if the Young Women cohort (age 14-24 in 1968, terminated as of 2003) could be re-interviewed at least one additional time, perhaps in 2021. If this were done, it would extend the Young Women cohort’s effective age range from 14-24 in 1968 to 67-77 in 2021. Assuming the NLSY79 continues as planned, a full, adult life-cycle of information would be available for women born in 1943-53 and women born in 1957-64. Thus, you would have two groups of women who have both gone through their adult working ages when dramatic marriage, employment, childbearing and subsequent marital transitions were occurring. With limited data manipulations, these two cohorts could be lined up to compare inputs and outcomes at ages 15-19 corresponding to adolescent events (schooling, early relationships and employment activities, and parental linkages/activities), ages 20-24, 25-29, 30-34, and so forth to 60-64. With additional data from the NLSY97, there will be the option of comparing three cohorts (Young Women, NLSY79, and NLSY97) born over a 40-year period from their teens into their 30s and beyond.

Cohort Birth
Age at 1st
Age at last
Age in
Mature Women 1922-37 30-44 (1967) 66-80 (2003)
Young Women 1943-53 14-24 (1968) 49-59 (2003) 67-77
NLSY79 1957-64 14-22 (1979) 50-58 (2015)† 56-64
NLSY97 1980-84 12-16 (1997) 29-33 (2014) † 36-40
†The NLSY97 was last interviewed in 2013-2014 and returns to the field in late 2015; the NLSY79 was last interviewed in 2014-15 and is scheduled for the next biennial interview in late 2016.

Adding the Mature Women cohort (age 30-44 in 1967, terminated as of 2003), a variety of analyses can be done that would include age-appropriate comparisons for more limited age ranges. Within the current NLS years, one can compare events using available data for women in the Mature Women, Young Women, and NLSY79 cohorts who are at some point between the ages of about 30 and 50. One can explore processes within each of these three cohorts, but in addition extend analyses to cross-cohort comparisons that include data for external factors such as employment conditions and quantifiable social transitions.

There are a significant number of common data elements that can be used as inputs and/or outcomes that can be compared across cohorts at different time points but at similar ages, with respect to a number of dynamic processes. This includes employment, family, marital transitions, and indeed a variety of attitudinal measures. For cohort analyses that begin at younger ages, in some instances, it is possible to include significant pre-marriage or relationship timelines to explore how pre-relationship youthful attitudes suggest independent longer term connections as well as indirect impacts through later behaviors. The options are considerable.

I have been very fortunate to have been heavily involved with the various NLS data sets over several decades. The NLS has grown decade by decade, including thousands of respondents from all parts of this country who volunteered time on a continuing basis— often for years—and provided unique and critical information that has permitted the completion of so many papers that have contributed to a deeper understanding of what has made our country function. This has included not only information about what we look like, but more importantly, the dynamics behind so many of our social, economic, and demographic behavioral paths. Partly because of NLS-based research, and how it has helped develop the analytical skill bases of so many as well as broaden our survey research capabilities, data collection and analyses are, I believe, far advanced from where they were 50 years ago.

However, I believe the NLS still has much to accomplish, partly by allowing ongoing cohorts to “complete” their lives. If we allow it to go on, perhaps even with additional cohorts, I believe our understanding of what “makes this country tick” will be enhanced, particularly by enhancing longer term understanding of the forces behind social change.

As a final point, regarding how highly I regard the NLS and its continuation, I will personalize this a bit more: I believe that ways in which the NLS program promoted its broad use for so long to scholars led to a personal spinoff—although it wasn’t always considered positively at the time. As with many disorganized social researchers, I frequently had piles of output all over the floor at home. We used lots of paper then! My son, perhaps in a particularly bored moment, picked some of the stuff up—and actually became interested in what it was. I explained what I was working on in a general sense, and he actually was interested. To make a long story short, he chose to use NLS data in his Ph.D. dissertation at the University of Illinois in 1996. This was maybe one of the earlier instances of an inter-generational transfer of the data set. Josh is neither a sociologist nor an economist but an epidemiologist—which is further testimony to the versatility of the NLS. Further, he has since included it on occasion in his research and, like father like son, he went on to promote the NLS at the Centers for Disease Control where he has spent his career.

… to the thousands of users of NLS data whose research produced roughly 9,000 journal articles, books, monographs, and dissertations during the first 50 years of the NLS. Without a dedicated community of NLS users, there would be no need for the NLS!

If you have used NLS data in your research and would like to add your “researcher” profile to this website, submit your profile information here.