The current study has implications for questioning young children regarding spatial descriptions of clothing. This is the bottom line of a recently published article in Law and Human Behavior. Below is a summary of the research and findings as well as a translation of this research into practice.
Featured Article | Law and Human Behavior | 2017, Vol. 41, No. 4, 398-409
Stacia N. Stolzenberg, Arizona State University
Kelly McWilliams and Thomas D. Lyon, University of Southern California
Children’s descriptions of clothing placement and touching with respect to clothing are central to assessing child sexual abuse allegations. This study examined children’s ability to answer the types of questions attorneys and interviewers typically ask about clothing, using the most common spatial terms
(on/off, outside/inside, over/under). Ninety-seven 3- to 6-year-olds were asked yes/no (e.g., “Is the shirt on?”), forced-choice (e.g., “Is the shirt on or off?”), open-choice (e.g., “Is the shirt on or off or something else?”), or where questions (e.g., “Where is the shirt?”) about clothing using a human figurine, clothing, and stickers. Across question types, children generally did well with simple clothing or sticker placement (e.g., pants completely on), except for yes/no questions about “over,” suggesting children had an underinclusive understanding of the word. When clothing or sticker placement was intermediate (e.g., pants around ankles, and therefore neither completely on nor off), children performed poorly except when asked where questions. A similar task using only stickers and boxes, analogous to forensic interviewers’ assessments of children’s understanding, was only weakly predictive of children’s ability to describe clothing. The results suggest that common methods of questioning young children about clothing may lead to substantial misinterpretation.
spatial language, investigative interviewing, child sexual abuse
Summary of the Research
“In investigations of alleged child sexual abuse, children are commonly asked to recall the placement of clothing or the nature of touching with relation to clothing. Such questions are often critical in assessing children’s allegations. Children’s responses may determine whether touching was abusive, and if so, the seriousness of the abuse. Most often, children are questioned with yes/no or forced-choice questions containing spatial terms (e.g., “Were your clothes on or off?”). Although developmental psychologists have mapped out children’s emerging understanding of spatial language, focusing on prepositions and verb phrases, and a great deal of applied psychology research has examined the effects of question type on children’s eyewitness performance, little is known about how well young children answer different types of questions incorporating spatial language. In practice, this is particularly important when there is a less than perfect fit between the situation and the spatial term. For example, if pants are around the ankles, how will children respond if asked if the pants are “on or off”?” (p. 398).
“We examined 3- to 6-year-old children’s ability to describe spatial relations in response to different types of questions (yes/no, forced-choice, open-choice, and where) using spatial terms that occur most frequently in questioning about sexual abuse: on/off, outside/inside, and over/under. We tested children’s understanding
in two tasks: one using stickers, clothing, and human figurines (the clothing task) and the other using stickers and boxes (the box task). The clothing task was an analog for children’s ability to describe clothing and touching with respect to clothing, and the box task was an analog for pretests assessing children’s ability to comprehend and use spatial language. We were particularly interested in children’s responses when presented with intermediate placement, that is, when clothing (or a sticker) was not completely on, outside, or over” (p. 400).
“The clothing task included the figurines, the shirt and shorts, and the stickers. RA1 placed clothing or stickers in various positions and RA2 asked the child to describe the placement. With respect to clothing placement, RA1 placed a shirt or shorts in five different placements: two simple placements (on or off) and three intermediate placements (clothing at a low joint: ankles/wrists; clothing at a midjoint: knees/elbows; clothing unfastened). The box task included the box and brightly colored stickers. The procedure was similar to the clothing task, with a sticker placed in three different placements for each spatial term: two simple placements (sticker on/off, outside/inside, over/under) and one intermediate placement (sticker partially on, sticker partially outside, sticker partially over)” (p. 401).
“Consistent with our predictions, we found that where questions were advantageous. When placement could be described with a single spatial term, where questions were as effective as other question types. When an intermediate description was more appropriate, where questions were clearly superior. We found that although the group patterns of responding to the box task were similar to that in the clothes task, individual children’s performance on the box task poorly predicted their performance on the clothes task. The practical implications of our results seem clear: When questioning young children, where questions appear superior in eliciting spatial descriptions of clothing than other types of questions, and pretests designed to assess understanding are only weakly diagnostic” (p. 406).
Translating Research into Practice
“Where questions were consistently superior to other question types in eliciting intermediate spatial descriptions from children. For example, when clothes were midjoint or low-joint (pants around knees or ankles, shirt around elbows or wrists), where questions elicited intermediate responses in 80% of children, whereas only 8% of the children asked where questions gave simple “on” responses. If asked yes/no questions, in contrast, children assented to simple “on” descriptions about half the time, simple “off” descriptions about 30% of the time, and provided intermediate descriptions only 10% of the time. Similar problems were observed with respect to forced-choice and open-choice questions, which often elicited simple responses when intermediate responses would be more accurate and informative. Furthermore, these problems were replicated when children were asked questions using outside/inside and over/under. Across tasks, where questions elicited intermediate descriptions in response to intermediate placement at least 44% of the time” (p. 406).
“Yes/no, forced-choice, and open-choice questions were thus quite likely to produce misleading responses. If clothes were only partially removed, then both “on” and “off” responses are likely to mislead a questioner; an “on” response would make some types of contact appear implausible, and an “off” response would make it appear that the child was describing complete rather than partial disrobing. In contrast, where questions were both less likely to elicit simple on and off responses and more likely to elicit a description that enables the interviewer to accurately envision clothing placement” (p. 406).
“An unexpected finding was that yes/no questions were difficult for children when they were asked questions using the word “over.” Children tended to deny that the sticker was “over” the clothing when the sticker was placed on the outer surface. Linguists have remarked on the large number of senses of “over,” noting that whereas some uses of “over” imply separation (e.g., “the plane flew over the house”), others imply contact (e.g., “the tablecloth was over the table”). Apparently, children interpreted “over” as implying separation, and thus denied that a sticker in contact with clothing was over the clothing. Children did not exhibit this difficulty when asked forced-choice questions about whether the sticker was “over or under” clothing; either the “or under” option provided context so that they understood the intended meaning of “over,” or they chose “over” simply because they recognized that “under” was clearly incorrect” (p. 406).
“Children’s difficulty with “over” highlights another benefit of where questions. Where questions enable children to choose spatial language with which they are most proficient. Indeed, when children were asked where questions in the conditions in which the other groups were asked over/under or outside/inside questions, they rarely used the terms “over” and “outside,” but preferred to say that the sticker was “on” the clothing or box. It is likely that similar difficulties will arise when children are asked yes/no questions using other spatial terms. For example, children who have not acquired the word “between” tend to use “in,” and children who have not acquired “above” tended to use “on.” If children are asked yes/no questions about “between” or “above,” they may well show patterns of errors similar to those found with respect to “over” in this study. Even if they find words totally incomprehensible, they are likely to attempt to answer closed-ended questions that contain those words” (p. 406).
“Practitioners are sometimes advised to assess children’s understanding of spatial terms before conducting their interview. However, researchers have had limited success in identifying useful pretests. For example, assessment of children’s truth/lie understanding has tended to find that it only weakly predicts children’s honesty when asked to promise to tell the truth. At first glance, the results of this study suggest that the box task had some utility: the pattern of results for performance on the box task was quite similar to that on the clothing task. The one notable exception was that children’s tendency to deny that stickers were “over” was less pronounced on the box task than on the clothing task, such that the difference on the box task was not statistically significant. Hence, had we only conducted the box task, we would have reached similar conclusions with respect to the utility of where questions for questioning children about clothing and touching with respect to clothing” (p. 407).
“However, analyzing individual children’s performance, we found that performance on the box task was only weakly diagnostic of their performance on the clothing task. This raises a general point about experimental results: A task can be a useful tool for understanding children’s performance as a group but a poor tool for assessing how individual children will perform. For example, when a group of children at a specific age perform well on a task, it may be better to assume that children at that age understand the task, rather than to individually test children. The test will offer little additional value, because the base rate of understanding in that age rate will be quite high. Conversely, if children are responding at chance, individual testing may have little diagnosticity, because a substantial proportion of children who pass are doing so randomly” (p. 407).
Other Interesting Tidbits for Researchers and Clinicians
“A topic for future research concerns the potentially negative effects of yes/no or forced-choice questioning on subsequent descriptions. If placement is intermediate, but a child is asked a closed-ended question and generates a response suggesting simple placement, how does this affect future interviews? If the child’s subsequent descriptions are intermediate, then the reports are inconsistent, which risks undermining the child’s credibility. If the child’s subsequent descriptions are simple, influenced by the initially closed-ended description, then the child’s report has been tainted” (p. 407).
“Furthermore, from a developmental perspective, our age range (3- to 6-year-olds) is quite large, and future work can elaborate on developmental differences. For example, larger samples with younger children may reveal difficulties in generating information required by the where questions, given younger children deficient recall abilities. Conversely, it is important to determine the extent to which even older children may fail to provide intermediate responses when interviewers ask closed-ended questions. For example, when shirts and pants are unfastened, even adults might exhibit a tendency to simply affirm that clothing is “on.” Larger samples (along with more extensive follow-up questioning) will also enable researchers to better understand other patterns of responding, such as double yes responses (e.g., “Yes” to both “Are the clothes off?” and “Are the clothes on?”). These responses might reflect response biases (and thus are likely to decrease with age), but they could also reflect sophisticated recognition of intermediate placement” (p. 407).
Join the Discussion
As always, please join the discussion below if you have thoughts or comments to add!