|The Double-Blind Lineup
General Comments and Observations
Gary L. Wells
A double-blind lineup is one in which the lineup administrator does not know which persons are fillers and which person is the suspected culprit. First introduced by Wells in 1988, the double-blind lineup is designed to prevent the person administering the lineup (traditionally the case detective) from inadvertently influencing the eyewitness’s identification decision or the certainty that the witness develops about the identification. Today, a number of jurisdictions across the US have adopted the double-blind lineup procedure, including the entire states of New Jersey and North Carolina, most of Wisconsin and Minnesota, and cities such as Boston, Denver and Dallas. Mistaken eyewitness identification is the leading cause of the conviction of innocent persons and I believe that this is the single most important reform that can be made to the way that lineups (both photographic and live) are conducted.
The double-blind lineup was first proposed in 19881, so the idea has been around for over 20 years. The double-blind lineup is simply a lineup that is conducted by someone who does not know which person in the lineup is the suspected culprit and which ones are mere fillers. I proposed this idea in 1988 because there is very good evidence that testers influence the person they test. The award-winning research of Robert Rosenthal and his colleagues has shown this type of influence in a broad array of situations over the years. The idea of a double-blind method of conducting lineups also follows closely on the compelling analogy between police conducting lineups with eyewitnesses and researchers conducting experiments with human participants2. Properly-conducted research studies use double-blind procedures (or some other method) to prevent the person who conducts the test from influencing the person they are testing. Double-blind testing is a core idea in science and medicine and the recommendation for double-blind lineups has been formally endorsed by the American Psychology-Law Society.3
Except for some jurisdictions that have reformed their eyewitness identification procedures (which I will mention later), the standard method of administering a lineup is to have the case detective administer it to the witness. Obviously, the case detective is fully aware of which person is the suspected person and which ones are merely lineup fillers. And, over 90% of initial eyewitness identifications are obtained using photo-lineups rather than live lineups. This observation about most lineups being done with photos is important because it means that there is no defense counsel present (the US Supreme Court has ruled that there is no right to defense counsel presence at photographic lineups). Furthermore, there is no video recording or audio recording of the interaction between the witness and the detective.
Since the first time I introduced the double-blind lineup in 1988, I have consistently maintained that the concern is not that investigators are intentionally (or knowingly) influencing witnesses, but rather that they are inadvertently (and often without any awareness) influencing witnesses. In general, people are unaware (or underestimate) the extent to which they influence the people they “test.”
Over the years, I have seen increasing amounts of evidence from actual criminal cases that have reinforced my firm belief that all lineups should be conducted using the double-blind method. The DNA exoneration cases that involved mistaken eyewitness identifications, which is the vast majority of all DNA-based exonerations, are a particularly interesting set of cases. In these cases, I have discovered numerous examples of the influence that the lineup administrators have on the eyewitnesses. In one case, the detective said to the witness while the witness was looking at a photo-lineup “I noticed you paused on number three.” In other cases, witnesses have uttered more than one photo number (e.g., “uh…it could be number four or it could be number two”) but the detective only follows up on the one who is the suspected person (e.g., “tell me about number four”) and ignores the one who was a filler. When witnesses have picked a filler, detectives have said “Is there any other photo that stands out?” which is a question that never gets asked when the witness picks the suspect. Remember, the suspect is just a suspected person; it is the person the detective had in mind when the lineup was created. There are now around 200 proven DNA exonerations involving mistaken identification from photo-lineups and live lineups and the innocent person who was convicted was the suspect in every one of those cases.
The idea of eyewitness identification evidence is that it is supposed to be based on the eyewitness’s memory for the perpetrator. Identification decisions should not be based on what the witness might glean from verbal or non-verbal cues from the case detective or anyone else. And yet, it is human nature for a person to unintentionally give off cues. Scientists have learned that the only way to prevent these types of interpersonal influence is to use double-blind techniques (or comparable procedures, such as having the testing be done by a computer rather than a person).
We have various lines of evidence from laboratory experiments on eyewitness identification showing that eyewitnesses are easily influenced by their testers. Some studies, for example, have led the persons conducting the lineups to believe (falsely) that a particular person in the lineup is the actual “culprit” and they simply switch around which person they lead the lineup administrator to think was the culprit. Then, the question is whether the eyewitnesses (to staged crimes) are more likely to pick the innocent person who the lineup administrator was led to believe did it. Sometimes this influence surfaces and sometimes it does not and it is difficult to predict when the effect will occur, but numerous published studies have found such effects4. The impact of even subtle nudges by a lineup administrator can be quite strong and, importantly, witnesses seem largely unaware of the fact that they have been influenced5. Particularly strong and consistent effects have been observed when lineup administrators react to witness’s identification decisions (e.g., confirming feedback such as “Good, you identified the suspect”). This type of confirming feedback after witnesses make mistaken identifications serves to dramatically inflate the confidence of the witness and make them believe that they were positive in their identification from the outset (when in fact they were uncertain). This effect was first demonstrated in 19886 but the literature has grown quite large to show that the effect is highly replicable7. These laboratory experiments have also been replicated in the field with actual eyewitnesses to serious crimes8.
The American Psychology-Law Society endorsed the idea of double-blind lineups in a “white paper” on the science of eyewitness identification and this was published in 19989. The state of New Jersey was the first state to require that lineups be conducted using double-blind methods (starting in 2001). More recently, North Carolina made a law requiring that all lineups be conducted using double-blind methods. Cities such as Minneapolis, Dallas, Denver, and Boston now use double-blind methods for lineups. But, by my estimates, three-fourths of the country still does not use double-blind lineups.
I have much more to say about the need for and value of double-blind lineups for helping to “clean up” eyewitness identification evidence. But, we must also keep in mind that double-blind lineups are not the whole solution to the eyewitness identification problem; instead, double-blind lineups are the foundation on which other reforms (such as how to best select lineup fillers, pre-lineup instructions, and so on).
RETURN TO WELLS HOME PAGE