Mcquiston and Malpass start out by giving a bit of background on what the Mockwitness Paradigm is and how it is intended to aid in the construction of proper eyewitness lineups. False positives are more likely to occur when a given suspect stands out in some way from the rest of the individuals in a particular lineup. This type of structural bias is supposedly addressed by the use of mockwitnesses.
The basic idea goes like this: Individuals who have not witnessed the crime are given the physical description of the perpetrator based on the description given by the actual eyewitness. These "mockwitnesses" are then asked to select an individual from a presented lineup based on the physically descriptive information they received. The point of this being, if the lineup is constructed properly and is structurally unbiased, each member of the lineup should have an equal chance of being selected and would not deviate from chance expectancy.
Mcquiston and Malpass admit to the sense of this method, but feel that the assumptions the mockwitness strategy rest on are theoretical and haven’t been subjected to empirical examinations. Three questions are asked
Question two addresses whether or not the type of instruction given to the mockwitness has an effect on the outcomes of their decisions. Mcquiston and Malpass are surprised at the lack of information on this topic as they feel that it certainly needs attention. Variance in the types of questions asked, leading vs. not, and manipulation of decision criterion, could confound the results if the variations have an effect.
Questions three touches on the assumption that supposedly the mockwitness are only using information given them to make their decision. M & M see a possiblilty of mockwitnesses bringing in their own prejudices and stereotypes as added information in making their evaluations.
Their goal in this research was to examine whether mockwitness’ evaluations for biases would vary based on changes in a given procedure. Secondly, they wanted to explore the source/s of mockwitness’ identifying information used in determining the target. They hypothesized that when mockwitnesses were provided with a structurally biased line up, they would choose targets in a manner significantly greater than chance expectancy for any one subject. In the same manner, given a structurally fair lineup, choice rates for targets would be about equal for each target, due to the similarity between them. Secondly, they hypothesized that mockwitnesses would make fewer errors when give unbiased/high criterion instructions vs. biases/low criterion instructions. And tossing in one more hypothesis for good measure, they expected that mockwitnesses would report that the physical description given to them would be their main source of identifying information.
Methods/Design/Results
Participants were 480 undergraduate students who were to serve as the mockwitnesses. The study was a 2 (lineup structure: fair vs. unfair) x 2 (linup instructions: biased vs. unbiased) x 2 (criterion instructions: high vs. low) x 2 (target one vs. target 2) between groups design with participants randomly assigned to each condition, 30 to a cell.
Four six-member lineups were created, two of which were fair (high target/filler similarity) and two that were structurally bias (medium target/filler similarity). The article goes into how these were put together in length; I’m sure we’ve all read it. Moving on.
Participants were given instructions including the physical description of a man who had recently committed an armed robbery, and a set of instructions depending on their experimental condition. The instruction sheet contained 20 numbered blanks for a "yes" or "no" decision as to whether or not each individual depicted in the presented photograph was the culprit. Participants were not made aware of how many pictures they would see, nor were they instructed to choose only one. They then viewed on of the four lineups sequentially with each picture being displayed for about 20 seconds. Lineup photographs were counterbalanced to prevent order effects. After the identification section, mockwitnesses were given a 4 item questionnaire that addressed decision processes involved in making their choices.
Results were out of the 480 participants, 403 made at least one choice. Of those, 167 made only one choice. The rest made multiple choices from 2 to all 6. Single choosers were placed in one catergory for analysis and multiple choosers in another.
A significant association was found between target identification accuracy and lineup structure in the single choice analysis. Eleven percent of the mockwitnesses who viewed a fair lineup correctly identified the target while 58 percent of those who viewed the unfair lineup identified the target.
For the multiple choosers, they were divided into accurate/inaccurate groupings. They were considered accurate if they got the target and inaccurate if they didn’t. Those that viewed a fair lineup were less likely to correctly identify the target. A significant interaction was found between Target Identification Accuracy x Lineup Structure x Lineup Instructions x Target. For the biased and unbiased instructions, unbiased in the unfair lineups made significantly more correct identifications for both target 1 and 2 than when the lineup was fair. A similar pattern was found on the biased instructions, but only with target 2.
Target position in the lineup was found to be a factor. Slot one was chosen at a rate that significantly deviated from chance. (I’m keeping this to two pages, by god)
Findings
Considering mockwitnesses that made
one choice, hypothesis one is supported by the data. Manipulations of the
instructions and criterion levels had no apparent effect. For the multiple
choice mockwitnesses, its less clear because both fair and unfair significantly
strayed from chance, but instructions and criterion level had no apparent
effect. The questionnaire showed that the main information used to decide
the culprit was the description given, although they noticed some participants
using stereotypes and schemas of "criminality" in their decision, something
that Mcquiston and Malpass deem worthy of further exploration.
|
|
|
|
|
|
|
|