printable version of page Printer-friendly page

Focus On Basics

Volume 1, Issue C ::: September 1997

Longitudinal Study of Adult Learners

by Barbara Garner
A longitudinal study is one in which researchers follow a group of people over time - in this case, for five years or more - periodically asking the same questions, looking for change. Different than comparing pre- and post-tests to judge program impact, longitudinal studies follow individuals and take into account a wide array of variables that make a difference in people's lives. Numerous small-scale, often in-depth surveys and ethnographic longitudinal studies of adult learners have been conducted in the past, providing the field with much needed knowledge. Missing, however, is a large-scale longitudinal study, one that includes enough learners to allow us to generalize findings to the population, and one that provides us with evidence of the impact of literacy learning over time.

Three Questions

With Steve Reder, Associate Professor of Psychology at Portland State University, as principle investigator, NCSALL hopes to follow a large group of adult learners: 5,000 to 15,000 might provide a nationally representative sample. Such large numbers are needed to make sure enough representatives of key subgroups - for example, Latinas - are included. The study will look at these broad research questions: 1) the growth of the study participants - literacy abilities along with other skills and knowledge; 2) the impact of various experiences, particularly participating in literacy education programs, on the development of their literacy abilities and other skills and knowledge; and 3) the relationship between improved literacy and other skills and knowledge on important personal, social, and economic aspects of participants' lives. This information will, in turn, let us say something substantive about the impact of adult literacy education, help us design more learner-centered systems, and add to our understanding of how soft' measures of adult learning - the kind of change teachers often see but find hard to document - are related to changes in standardized test scores.

Design Challenges

The first question to answer in designing a longitudinal study is "What questions are we hoping to answer?" Through an iterative process involving many policymakers and practitioners, Reder has decided upon the three questions listed above. Next, he must decide whom to include in the survey - the range of skill levels and ages to include are just two of the decisions to be made. "The biggest question is whether to include the ESOL (English for Speakers of Other Languages) population or not. It means lots of extra work in terms of instruments, procedures, cross validation issues...," says Reder. In other words, designing instruments and questions that measure the same things in the same way across languages is difficult. And then we must ask, "Which other languages?" For each additional language, a large enough subgroup must be surveyed to provide reliable information, additional surveyors must be recruited and trained, and field procedures must be monitored to assure that one everyone's work is consistent. This quickly inflates the size, complexity, and cost of the study. Although almost everyone wants to include ESOL learners in the study, it will be challenging to secure the extra resources needed to do so. At publication date, NCSALL is still exploring the costs and tradeoffs among various study designs and target populations.

Since longitudinal studies are costly, NCSALL must develop cost-effective techniques for collecting needed information from participants. A major issue is keeping track of study participants between rounds - or waves'- of data collection, so that people can be inexpensively located if they move. Another issue is collecting data face-to-face as opposed to over the telephone. The initial wave will likely include questions asked and literacy assessments conducted in person, with survey staff visiting participants in their homes. Subsequent waves could be collected via telephone at a considerable cost savings compared to face-to-face interviews. But, to use telephone interviews effectively in this study, the challenge of gathering accurate information about adult learner's literacy proficiencies over the phone must be met. Reder is considering pilot-testing innovative ways to assess literacy. Whether some combination of unobtrusive methods will validly measure literacy gains over time is a question Reder and his colleagues will be looking at as they design the study.

An additional design challenge is deciding which of many possible topics to ask questions about, and how to ask them so that the answers are accurate and useful. Once survey questions are decided upon, the survey will be drafted and tested, revised and retested. At each step of the process, Reder will solicit input from members of the adult literacy field.

For updates on the longitudinal study, return to the NCSALL web page.

or contact:

Steve Reder
Portland State University
P.O. Box 751
Portland, OR 97207
phone: (503) 725-3999
fax: (503) 725-3904
and e-mail: reders@pdx.edu.

Updated 7/27/07 :: Copyright © 2005 NCSALL