Editor’s note: This essay is part of a series being published to help commemorate the Monthly Labor Review’s centennial (July 1915–July 2015). The essays―written by eminent authorities and distinguished experts in a broad range of fields―cover a variety of topics pertinent to the Review and the work of the Bureau of Labor Statistics. Each essay is unique and comprises the words and opinion of the author. We’ve found these essays to be enlightening and inspirational. We hope you do as well.
For most of the Monthly Labor Review’s 100-year history, sample surveys have been a critical source of data for reporting trends in employment and consumer behaviors. The data from these surveys have provided a much needed understanding of the performance of, and trends in, the nation’s economy. In the future, it is likely that administrative records and the constant monitoring of the electronic footprints left by people and businesses in their daily activities will replace the need to ask certain survey questions.
However, it is not likely that these new sources of information will mostly replace the collection of critical data through periodic sample surveys. The unfolding of new economic challenges related to work and to consumer issues will continue to require the asking of survey questions on new topics in order to obtain information critical to maintaining and growing our national economy.
Yet, our current modes of survey data collection are increasingly inadequate. Response rates to voluntary telephone and business surveys have declined, in some cases precipitously. Random-digit dialing (RDD) telephone surveys, the main workhorse for the last 30 years of household surveys, now have severe coverage issues as well as response problems. Internet surveys, the much anticipated replacement for RDD has no email sample frame comparable to the RDD procedure that facilitated probability household surveys by telephone. Although Internet access exists for the vast majority of households, getting responses over the Internet from email contact is proving as difficult to achieve as getting people to answer telephones. The sheer volume of email messages, and the use of multiple devices to assist with their management, has ushered in a swipe-and-delete attitude that leads to poor response rates, in the same way that answering machines contributed to survey requests being blocked or ignored.
Fortunately, single-mode surveys are being replaced by mixed-mode designs that use multiple contact as well as the offering of different response modes to garner people’s attention and interest in responding to survey requests. Being able to contact people sequentially by email, text, voice telephone, and postal mail and delivery services makes it possible to reach individuals and businesses more often and with greater effectiveness. Multiple forms of contact facilitate building synergy across contact modes, increasing the likelihood that more-effective requests for explaining survey needs and obtaining responses can be delivered. A proven example is email augmentation, whereby an initial postal contact to legitimize a study is followed immediately by an email that provides an electronic link to the survey. That combination of contacts provides a win–win situation: it is easier for the sample unit to respond in a timely way, and the quality of the data goes up from higher response rates. In addition, offering the possibility for respondents to answer by different modes improves the likelihood of obtaining representative responses.
However, combining responses from different modes—particularly if one mode is aural and the other(s) visual—introduces measurement differences. Research has shown that such differences can be reduced through unifying how questions are asked across modes. Instead of maximizing design for a particular mode, questions serving as stimuli now need to optimize in ways that make them the same across modes. In addition, nearly 20 years of research has shown that people’s answers to visual survey questions are communicated by numbers, symbols, and graphics as well as words, all of which must be taken into account in survey design.
In addition, the increasing amounts of times spent on portable devices and tablets that utilize smaller screens are moving us into an era in which space-rich question formats, such as matrix questions or a series of items asked as a group, e.g. how frequently different forms of transportation are used in going to and from work, need to be constructed differently than has traditionally been done for computer and paper surveys. As the use of the many dozen portable devices with smaller screens becomes more important for answering surveys, how we ask certain kinds of questions also must be rethought.
The emergence of mixed-mode data collection is making good survey design more challenging. Survey designers are necessarily relearning their craft so that they think multimode first. It is essential that the use of the telephone, mail, the World Wide Web, and the increasingly prevalent mobile devices be integrated effectively to achieve a good response. The quality of sample survey results made available to users through the Monthly Labor Review depends upon realizing this transition.
Don A. Dillman, "Future surveys," Monthly Labor Review, U.S. Bureau of Labor Statistics, November 2015, https://doi.org/10.21916/mlr.2015.44.