Abstract
Despite over 40 years of study, question order is probably the least developed and most problematic aspect of survey research. As Schuman and Presser (1981) remarked in their work on survey methodology:
Overall, order effects ⋯ constitute one of the most important areas for methodological research. They can be very large [and] are difficult to predict⋯. At this point research needs to be aimed not merely at producing more examples, but at understanding why those already obtained occur, (p. 77).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
A result that is at odds with Schuman and Presser’s similar experiment (1981; Schuman, Presser, & Ludwig, 1981).
The Communist/American reporters example is actually more complicated than the others because the marginal effects are reciprocal. As a result, the distribution of the conditional controls varies by order.
These results are contrary to those reported by these authors. They generally report no such conditional effects. For example, “Similarly, the effects of context did not depend on the respondent’s initial opinion about the target issue” (Tourangeau et al., 1988, p. 30).
Looking at the six abortion items on the GSS, which include the two items used by Schuman and Presser, we find that the general abortion item is the hardest item to approve, while the birth defect item is the second easiest. The coefficients of reproducibility and scalability are.94 and.81.
In terms of the cognitive framework of Tourangeau and Rasinski that we explore later, there are several ways that conditional order effects could be created. Even if no relevant attitude was directly expressed, a retrieval carryover effect could occur if a prior question triggered selective memory sampling and people differed in their affect toward the primed memories. In addition, the expression of a relevant prior attitude could create a conditional order effect by causing judgmental carryover or consistency editing during the response selection stage.
Also see Bradburn and Mason (1964), which tested for 14 differences in marginals across four forms and found no statistically significant variation.
The abortion example discussed above is a prime example of what can occur when the order within such a block is disturbed. See also Schuman and Presser’s (1981) discussion of same. For other examples, see Astin et al. (1988) and T. W. Smith (1984). I suspect that within-scale context effects are rather common, probably even typical. Since such scales tend to be replicated as units without changes to their internal order, these effects are rarely studied, and whatever context effects exist generally remain fixed across administrations of the scale.
Here, as elsewhere in this chapter, we exclude the related matter of response-order effects (see Schuman & Presser, 1981, pp. 56–74).
The authors sometimes refer to these as assimilation and contrast, respectively.
The basic categories are employed, but (a) some additional refinements and distinctions are added, along with some new terminology, and (b) the part-whole distinction is not utilized (T. W. Smith, 1986a).
But for an exception, see Bishop, Oldendick, & Tuchfarber (1982, 1984a).
For example, where context failed to induce logical constraint, see T. W. Smith (1981a, 1981c).
I found little support for this particular example (T. W. Smith, 1983a).
Tourangeau and Rasinski (1988) explicitly admit that their framework is not comprehensive and separately discuss sequence effects.
As Tourangeau and Rasinski (1988) note, “Context effects are often unstable; this instability may reflect the number and complexity of the processes that are responsible for the effects, as well as the large number of variables that can influence the size and direction of the context effects” (p. 311).
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1992 Springer-Verlag New York, Inc.
About this chapter
Cite this chapter
Smith, T.W. (1992). Thoughts on the Nature of Context Effects. In: Schwarz, N., Sudman, S. (eds) Context Effects in Social and Psychological Research. Springer, New York, NY. https://doi.org/10.1007/978-1-4612-2848-6_12
Download citation
DOI: https://doi.org/10.1007/978-1-4612-2848-6_12
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4612-7695-1
Online ISBN: 978-1-4612-2848-6
eBook Packages: Springer Book Archive