Eyewitness identification in simultaneous and sequential lineups: an investigation of position effects using receiver operating characteristics.

For decades, sequential lineups have been considered superior to simultaneous lineups in the context of eyewitness identification. However, most of the research leading to this conclusion was based on the analysis of diagnosticity ratios that do not control for the respondent's response criterion. Recent research based on the analysis of ROC curves has found either equal discriminability for sequential and simultaneous lineups, or higher discriminability for simultaneous lineups. Some evidence for potential position effects and for criterion shifts in sequential lineups has also been reported. Using ROC curve analysis, we investigated the effects of the suspect's position on discriminability and response criteria in both simultaneous and sequential lineups. We found that sequential lineups suffered from an unwanted position effect. Respondents employed a strict criterion for the earliest lineup positions, and shifted to a more liberal criterion for later positions. No position effects and no criterion shifts were observed in simultaneous lineups. This result suggests that sequential lineups are not superior to simultaneous lineups, and may give rise to unwanted position effects that have to be considered when conducting police lineups.

PageFocus: Using paradata to detect and prevent cheating on online achievement tests.

Cheating threatens the validity of unproctored online achievement tests. To address this problem, we developed PageFocus, a JavaScript that detects when participants abandon test pages by switching to another window or browser tab. In a first study, we aimed at testing whether PageFocus could detect and prevent cheating. We asked 115 lab and 186 online participants to complete a knowledge test comprising items that were difficult to answer but easy to look up on the Internet. Half of the participants were invited to look up the solutions, which significantly increased their test scores. The PageFocus script detected test takers who abandoned the test page with very high sensitivity and specificity, and successfully reduced cheating by generating a popup message that asked participants not to cheat. In a second study, 510 online participants completed a knowledge test comprising items that could easily be looked up and a reasoning task involving matrices that were impossible to look up. In a first group, a performance-related monetary reward was promised to the top scorers; in a second group, participants took part in a lottery that provided performance-unrelated rewards; and in a third group, no incentive was offered. PageFocus revealed that participants cheated more when performance-related incentives were offered. As expected, however, this effect was limited to items that could easily be looked up. We recommend that PageFocus be routinely employed to detect and prevent cheating on online achievement tests.

Seriousness checks are useful to improve data validity in online research.

Nonserious answering behavior increases noise and reduces experimental power; it is therefore one of the most important threats to the validity of online research. A simple way to address the problem is to ask respondents about the seriousness of their participation and to exclude self-declared nonserious participants from analysis. To validate this approach, a survey was conducted in the week prior to the German 2009 federal election to the Bundestag. Serious participants answered a number of attitudinal and behavioral questions in a more consistent and predictively valid manner than did nonserious participants. We therefore recommend routinely employing seriousness checks in online surveys to improve data validity.

cocor: a comprehensive solution for the statistical comparison of correlations.

A valid comparison of the magnitude of two correlations requires researchers to directly contrast the correlations using an appropriate statistical test. In many popular statistics packages, however, tests for the significance of the difference between correlations are missing. To close this gap, we introduce cocor, a free software package for the R programming language. The cocor package covers a broad range of tests including the comparisons of independent and dependent correlations with either overlapping or nonoverlapping variables. The package also includes an implementation of Zou's confidence interval for all of these comparisons. The platform independent cocor package enhances the R statistical computing environment and is available for scripting. Two different graphical user interfaces-a plugin for RKWard and a web interface-make cocor a convenient and user-friendly tool.

Looking back on the London Olympics: Independent outcome and hindsight effects in decision evaluation.

Outcome bias and hindsight bias are related, but how exactly? To remedy theoretical ambiguity and non-existent directly relevant empirical research, we contrast an older idea (Baron & Hershey, 1988, J. Pers. Soc. Psychol., 54, 569) that sees outcome bias as partly mediated through hindsight bias with the idea that the two biases independently affect decision evaluations. In an Internet study of retrospections on the 2012 London Olympics, evaluations of the Games' success and its foreseeability had independent effects on evaluations of the International Olympic Committee's decision to award the Olympics to London; there was no evidence of mediation. Further theoretical discussion emphasizes the need to distinguish between a holistic assessment of decisions and a more specific assessment of the decision-making process in future outcome bias research.

A Strong Validation of the Crosswise Model Using Experimentally-Induced Cheating Behavior.

We constructed an online cheating paradigm that could be used to validate the Crosswise Model ( Yu, Tian, & Tang, 2008 ), a promising indirect questioning technique designed to control for socially desirable responding on sensitive questions. Participants qualified for a reward only if they could identify the target words from three anagrams, one of which was virtually unsolvable as shown on a pretest. Of the 664 participants, 15.5% overreported their performance and were categorized as cheaters. When participants were asked to report whether they had cheated, a conventional direct question resulted in a substantial underestimate (5.1%) of the known prevalence of cheaters. Using a CWM question resulted in a more accurate estimate (13.0%). This result shows that the CWM can be used to control for socially desirable responding and provides estimates that are much closer to the known prevalence of a sensitive personal attribute than those obtained using a direct question.