This is a quick FYI email-based post. Unfortunately, this is not an open access article 😕.
Abstract
With the growing use of touchscreen devices in cognitive and academic testing, understanding the impact of tap latency is important for ensuring test fairness, particularly as related to speed tasks. The present study aims to understand how tap latency influences participant test-taking behaviors and performance. Using a counterbalanced within-subjects design, 203 participants aged 5–19 completed three WJ V speeded subtests on both low-latency normative administration devices (i.e., iPads) and Android tablets with an experimentally imposed, noticeable 340-ms tap latency. While the scores achieved across the two different devices were generally consistent, the actual Android scores were significantly higher than scores predicted based solely on latency-related time loss across all tasks, suggesting behavioral compensation from the tap delay. While an argument can be made that scores are comparable and thus acceptable, given tap latency’s behavioral effects and the absence of validated post-hoc score correction models, it is recommended that WJ V speeded tests be conducted on devices with minimal and consistent latency and devices with unknown, variable, or consistently higher than 340-ms tap latency should be used with caution for speeded testing.
Some select article quotes (and my comments) from the article:
The present findings also provide evidence for the behavioral impact of tap latency, such that participants tested on high-latency tablets adapted their behavior in ways that mitigated its effects.
Therefore, many non-iPad devices may exhibit latencies larger than 340 ms, which could result in more pronounced behavioral differences in these tasks. As such, we recommend that WJ V speeded tests be administered on devices with low and stable tap latency, ideally iPads given its minimal tap latency and its use in the normative sample…
My comment: By changing their cognitive strategies, which was evidenced by adaptation in their behavior, the non-standard iPad high latency Android condition introduced construct irrelevant variance into the subjects scores—that is, by changing their strategies to compensate for slower latencies, the subjects changed what the speeded test was measuring…a threat to construct validity. The authors recommend the use of the iPad’s noted in the WJ V Technical Manual (LaForte, Dailey, & McGrew, 2025).
When administering the WJ V, Riverside Insights recommends the examinee device to be an iPad with a screen size of 10” or larger, as that is how the test was standardized. Because the assessment is browser-based, we recognize the examinee can be on any tablet touchscreen device with a screen of 10” or larger and Riverside Insights cannot control the device used. Please note, using a tablet device other than an iPad on timed tests may result in differences in scores, based on the latency times of different devices. Riverside Insights strongly encourages the use of an iPad, especially on timed tests.
COI statement: I am a coauthor of the WJ V, as well as prior editions (WJ III and WJ V). However, all WJ V authors no longer receive traditional sales-based royalties—payment for authoring services occurred prior to the release of the WJ V. My more complete COI statement is available here.
