Search Type Title Author/Investigator Search Advanced Search

Table of Contents | Previous | Next

Potential Opportunities within the NLSY97 and Next Steps

A number of potential opportunities for enhancement were recommended during the workshop that could significantly improve the ability of the NLSY97 to inform marriage and family research. These include several that would enable researchers and policy makers to more easily tap the NLSY97 as a ready source of information on how today’s families are doing. Other changes or new efforts would form a platform for greater collaboration on marriage and family research across policy areas and organizations, and establish a common body of knowledge that could help leverage advancements in the field. Finally, the panel suggested efforts they felt were vital to support movement beyond our current understanding of family functioning and prepare for the key family research questions on the horizon and the next generation of families.

In the table below and the discussion that follows, we follow up on the panel’s recommendations to provide a general sense of the efforts one would need to pursue these types objectives. Our description of potential next steps is suggestive, meant to serve as a platform for those with direct responsibility as well as members of the NLSY-user community to consider what range of efforts may be well-suited to their mission and means, rather than as a definitive proposal to complete such tasks.

Many of the suggested efforts would need to be explored in partnership with BLS, the sponsoring agency, particularly those involving content or delivery of the data. Others, such as targeted documentation, could go forward independently, although BLS would still be a valuable partner. These caveats notwithstanding, the illustrations below suggest the data activities agencies or organizations interested in family research and policy might wish to consider, and the steps that could take them in that direction. We discuss each of the major categories shown in the table, and the items within them, in turn.

Potential Opportunities to Increase the Utility of the NLSY97 Data

	Opportunity	Approx. Timing	Type of Effort	Possible Approaches
Transform NLSY97 into Ready Reference on Marriage and Family	Create targeted users’ guides on marriage and family topics.	Under 12 months. Could be initiated at any time.	Small effort for each topic	Pursue through BLS and its contractors, or convene group of data and subject experts.
	Create on-line table generator of family variables and covariates.	Under 12 months. Could be initiated at any time.	Medium	Attach to WebInvestigator, through BLS contract, or create alternative web site.
	Construct flat file of key marriage and family outcomes w/covariates.	12 to 24 months.	Medium	Pursue through BLS and its contractors, or convene group of data and subject experts.
	Assess sample adequacy for key populations.	Under 12 months. Could be initiated at any time.	Small	Could be done by any NLSY97 data expert. Suggest getting input from BLS or its contractors.
	Provide additional created variables to identify key relationships.	Under 12 months. Could be initiated at any time. Need to start 6 months prior to next data release.	Small effort for each set of variables.	Users can readily create and share variables on their own. Or variables can be developed and added to official data by BLS and its contractors.
	Define universe of eligible respondents by variable.	Computer Assisted Phone Interview Feasibility study: under 12 months. Implementation: up to 24 months.	CAPI Feasibility Study: Small Implementation:Large	Expert data users could do manually for select variables. Theoretically could be done globally as CAPI software enhancement through BLS and its contractors.
Create Common Knowledge Base and Platform for Collaboration/Leveraging	Create on-line collaboratory or message board.	Under 12 months to start-up. Could be initiated at any time.	Small effort to start up	Link to NLSY web through BLS and its contractors, or create alternative forums.
	Create on-line repository for code, tabulations, etc.	Under 12 months to start-up. Could be initiated at any time.	Small effort to start up	Work through BLS and its contractors to add to NLSY website, or create new site.
	Perform analyses of key topics using available data.	Ongoing, but key questions remain unanswered.	Small	Fund work through special journal issues, conferences, dissertation grants, etc.
Deepen Understanding of Families and Plan for the Next Generation of Research	Add key survey questions.	Design work must start 2 years prior to next data wave.	Variable	Proposals can be sent to NLSY97 PI. Must be done through BLS contract.
	Interview other people in respondents’ lives (e.g., partners, children).	Must select roughly 3 years prior to fielding.	Large	Proposals can be sent to NLSY97 PI. Must be done through BLS contract.
	Add new respondents (e.g., immigrant refresher sample).	Must select roughly 3 years prior to fielding.	Large to very large	Proposals can be sent to NLSY97 PI. Must be done through BLS contract.
	Create new cohorts, e.g. NLSY2010 or Children of the NLSY97.	Roughly 5 years from agreement to data collection.	Very large	Proposals can be sent to NLSY97 PI. Must be done through BLS contract.

Transform the NLSY97 into Ready Reference on Marriage and Family

Create targeted users’ guides on marriage and family topics

The NLSY97 Users’ Guide is an extensive document that not only describes the critical elements of data structure, sampling design, weighting and fielding procedures, but also provides overviews of topical content in the areas of: aptitude, achievement and intelligence scores; education; employment; event history; expectations, attitudes, behaviors, and time use; family background and demographic characteristics; health; income, program participation and assets; marital history and fertility; and training. This breadth of content limits the detail that can be presented on any one set of outcomes. At the same time, variables pertaining to topics such as relationship pathways and fatherhood number in the thousands and cut across topical modules, making it difficult to fully exploit the currently available data.

Targeted users’ guides could present the NLSY97 data from the perspective of marriage and family researchers. Guides could be developed, for example, pertaining to fatherhood, step-parenting, teen parenthood, healthy marriage characterization, or a variety of other topics. Content could include a more extensive discussion of relevant variables, a clearer delineation of variables by generation, caveats pertaining to particular measures or subgroups and information about missing values, control totals for key subgroups as well as the software code used to identify them, and cross references to other outcomes of importance, such as employment or health. As one possible first step in this process, a working group could be convened consisting of subject matter and data experts to discuss the topical and technical content.

Create On-Line Table Generator of Family Variables and Covariates

BLS provides extensive on-line support with respect to downloading the NLSY data and documentation, which is highly useful for those researchers undertaking the complex analyses that NLSY97 can support. However, some important and highly visible questions in the area of marriage and the family are primarily descriptive. People in the research and policy communities looking for quick and simple tabulations may avoid using the NLSY97 because the effort to understand the data and identify variables of interest is prohibitive for them.

One approach that could add significant currency and visibility to the NLSY97 data would be development of an on-line table generator similar to that provided by the Census Bureau for several of its data sets. Content could include not only outcomes related to marriage and family formation, but also the intersections of these variables with other important domains such as employment, education and health. A possible first step in this process could be the convening of subject matter and data experts to discuss the topical and technical content, as mentioned above.

Provide Additional Created Variables

Providing a number of identifying variables, based on existing data, could be particularly helpful in clarifying relationships, such as whether the respondent’s cohabiting partner is a biological parent of her child, and the type of relationship a respondent has with the children’s other parent. Suggestions on which new variables to construct could be solicited from experienced NLSY97 users, possibly through the on-line collaboratory or message board suggested below.

Define Universe of Eligible Respondents by Variable

On-line documentation for each variable specifies the number of valid and invalid skips in the data, as well as the survey questions leading into and following from that variable. While this provides essential information for assessing the quality of a particular measure, interpreting the meaning of valid skips remains challenging. If, for example, 80 percent of a researcher’s sample of nonresident fathers shows a valid skip for the amount of child support paid, this skip may arise either because those nonresident fathers already indicated elsewhere that they paid none, or because the researchers incorrectly identified the sample of nonresident fathers. To know for sure, one must trace through a sometimes lengthy list of preceding variables to untangle skip patterns. This is due to the fact that the pathway of questioning is highly tailored to each respondent’s circumstances through a highly complex CAPI instrument.

One enhancement that would make the interpretation of valid skips more transparent would be inclusion of a definitional statement for each variable identifying the universe of eligible respondents. This could be done manually by any knowledgeable data user, with an estimated time of 10 minutes per variable. Given a list of variables totaling over 20,000, a particular user would likely undertake this task only for a variable subset of particular interest. Alternatively, BLS and its contractors could enhance the CAPI program such that this information is produced globally for all variables. While this appears doable in theory, additional work would be needed to assess more closely the feasibility and level of effort needed.

Create Common Knowledge Base and Platform for Collaboration/Leveraging

Create On-Line Collaboratory or Message Board

Another type of on-line resource that could greatly enhance the usability and visibility of the NLSY97 is a message board, collaboratory or other on-line discussion forum dedicated to uses of the NLSY97 data for research on marriage and family. This forum would enable and encourage researchers to post findings, announce events, discuss methodology, and contribute to the planning of future NLSY enhancements. The forum could also be used to discuss difficult constructs or share knowledge and encourage family research that extends across policy domains. Such a forum could be pursued either through BLS and its contractors, or independently. Either way, one possible first step would be to convene a working group of subject matter and technical experts to discuss needs and structure.

Create On-Line Repository for Code, Tabulations, Measures, etc.

Similarly, one could create an on-line repository of computer code, tabulations, and definitions of key measures or subpopulations related to NLSY97 research on marriage and the family. Such a repository would make important advances in the field available to a wider spectrum of researchers, and thus facilitate greater progress on important unanswered questions pertaining to relationships and family well-being.

Promote Analyses of Key Topics Using Available Data

Analyses on a number of key topics using the data as they currently exist also could be instrumental in making the utility of these data more visible. One such topic suggested by the panel is the extent to which marriage outcomes are driven more by opportunities in the marriage market or personal values related to marriage, and the ramifications of that for programs or policy. Research on these and other topics of interest could be encouraged through the provision of small dissertation grants or other funding opportunities. Such steps could be pursued through a wide range of avenues and organizations.

Deepen Understanding of Families and Planning for Next Generation

Add Key Survey Questions

While the NLSY97 data in its current state offers important opportunities to study family relationships and well-being, the panel felt that in order to pursue the important topics on the forefront of the field, new survey questions are needed on topics such as:

• relationship skills
• resource and power sharing
• balancing work and family
• financial literacy
• direct questions on the reasons for not engaging in particular events (e.g., working, getting married, paying child support or visiting nonresident child)
• direct questions on the timing and outcomes associated with key transition events
• more measures of co-parenting and the role of extended family
• in-kind child support provided by nonresident parent of R’s child
• perceived exit costs regarding current relationship
• trust of partner or worries of infidelity
• participation in or awareness of healthy marriage services
• occurrences of domestic violence
• the distinction between cohabitors who clearly see themselves on a path toward marriage versus those who do not
• outcomes of the focal respondent’s child
• a more extensive rostering of child/parental relationships inside and outside the household.

Changes to the survey can only be pursued through BLS and its contractors, and work would need to begin well in advance of the next wave of data collection. The cost of adding a question depends on the complexity of the question proposed and the amount of development and CAPI programming that would be needed. Since space on the NLSY97 survey is limited, and proposals for new questions would have to be evaluated within that context.

Interview Additional People in Respondents’ Lives

Panel members felt it was important to interview beyond the focal respondent for two primary reasons. First, they felt that capturing spheres of influence beyond the household is critical to understanding child wellbeing, given the increasing likelihood that children spend significant amounts of time being cared for in other households. Additional interviewees could include nonresident parents, partners or extended family members. Second, the panel believed that current measures such as relationship quality or parental engagement would be more reliable if gathered from both parties involved. In addition, panel members suggested that some questions could also be asked of the focal respondents’ children once they are older. Proposals for such expansions could be sent to the PI of the NLSY97, and would need to be approved by BLS.

Add New Respondents to Sample

The panel suggested that the current set of respondents be refreshed with a supplemental immigrant sample to make it more representative of today’s young adults. They cited important differences in family formation, employment and other outcomes among immigrants, their growing significance in the population, and the policy relevance of these dynamics. Major changes to the sampling frame can only be pursued through BLS and its contractors, require significant lead time in advance of the next wave of data.

Create New Cohorts

The most extensive steps the panel discussed pertain to development of new cohorts within the NLS program. New cohorts could represent children born to mothers in the NLSY97, or an entirely new cross section of youth in 2010. The 2010 option is more likely and, as noted, is already the topic of serious discussions between BLS and other government agencies including the Department of Defense. Organizations interested in the content of future NLSY cohorts should become engaged in such discussions well in advance.

 

 

Table of Contents | Previous | Next