Are identity lineups fair?

By Laurie Scarborough

The use of identification lineups, has been well-established and they are often the most important evidence against a suspect (Wells, Leippe, & Ostrom, 1979). Despite this, there are many difficulties with identification parades (Ainsworth, 1998; Wells et al., 1979). Lineup fairness and bias, in terms of procedure and construction, is a concern. Not only that but witnesses have to remember the face of a culprit they may have seen weeks before, or only seen briefly (Tredoux & Chiroro, 2005). Because of these challenges, lineups have led to hundreds of known false identifications resulting in wrongful incarcerations (Brigham & Pfeifer, 1994; Davies & Griffiths, 2008; Smith, Lindsay, & Pryke, 2000). In fact, false identifications are the leading cause of wrongful conviction (Naude, 2015; Tredoux & Chiroro, 2005). In many cases, wrongful convictions have been confirmed after DNA exonerations (Malpass, Tredoux, & McQuiston-Surret, 2007; Tredoux, 2011). In numerous instances the witnesses have made honest but wrongful identifications, but on other occasions misidentifications were the result of unfair or biased parades.

An example of a wrongful conviction from a biased lineup is the case of Adolf Beck. He was convicted of fraud, based on 12 identifications at a lineup, and sentence to six years in prison (Davies & Griffiths, 2008; Tredoux, 2011; Tredoux & Chiroro, 2005). He was later rearrested for the same crime before authorities apprehended the real perpetrator, Wilhelm Meyer, who bore a close resemblance to Beck (Tredoux, 2011). The lineup was biased against Beck because he was the only member of the parade with a moustache, one of the key features of the suspect’s description (Davies, & Griffiths, 2008). Misidentifications like that could be avoided if the lineup is fair (Malpass et al., 2007). In this case if the suspect was not distinctive in the lineup, the lineup construction would have been fairer.

A lineup is considered fair when mock witnesses choose each lineup member in equal proportion expected by chance (Ross, Tredoux, & Malpass, 2014; Tredoux, 1998; Wells et al., 1979), or biased if a lineup member is chosen more frequently than expected by chance (Ross et al., 2014; Tredoux, 1998; Wells et al., 1979), or if a mock witness can identify the suspect based on the suspect’s description alone (Malpass & Lindsay, 1999). Lineup fairness is affected by system variables, which are factors that can be controlled by the legal system (Wells et al., 1979). Unfortunately in South Africa, police officers do not have compulsory training for the administration of lineups, and the optional training they do receive is brief (Tredoux & Chiroro, 2005), so their expertise in handling system variables is sometimes minimal. These system variables are factors such as the number of lineup members, whether foils (non-suspects) in the lineup and the suspect look similar, and whether there is any suggestion to witnesses about the identity of the suspect from the lineup administrative office (Wells et al., 1979; Smith et al., 2000). These system variables will be explored in terms of lineup fairness in the following discussion.


Foil similarity to suspect

Paragraph 11(2) of South Africa’s National Instruction 1 of 2007 states that foils in a lineup must look similar to the suspect (Naude, 2015). Other sources specify that foils and suspects must be similar in terms of age, height, body type, race, hair, demeanour, position in life and attire (Ainsworth, 1998; Clarke, 2012; Davies & Griffiths, 2008; Malpass & Devine, 1985; Malpass, et al., 2007; Tredoux & Chiroro, 2005). No one should stand out of a lineup to suggest who is or is not a suspect (Clarke, 2012; Malpass & Lindsay, 1999; Malpass, et al., 2007; US Department of Justice, 1999). Lineups with similar-looking suspects and foils result in less false identifications, reducing the chance of this from 70% to 31% (Clarke, 2012). Thus, if foils are not similar to the suspect, and suspects are distinctive in a lineup, they are more likely to be chosen (Wells et al., 1979), resulting in a suggestive and biased lineup (Lindsay, 1994; Malpass & Devine, 1985). One such example is the robbery case of Segomotse v S. Witnesses described the culprit as having dreadlocks, but the only member of the lineup with dreadlocks was the suspect, therefore only the suspect matched the culprit’s description (Moore, J.J.A., McNally, J.J.A., Foxcroft J.J.A. in Segomotso v S).

However, foils should not be so similar to the suspect that is it difficult to distinguish between them (Brigham & Pfeifer, 1994; Malpass, et al., 2007; Naude, 2015; US Department of Justice, 1999), as then the suspect becomes the most familiar face (a prototype face) and becomes distinctive in that way (Marwitz & Wogalter, 1988). This increases misidentifications (Brigham & Pfeifer, 1994). Brigham and Pfiefer (1994) thus suggest that foils differ from the suspect in features that were not mentioned in the suspect’s description.

There is also debate around whether foils should be selected based on their similarity to the suspect or based on their suitability to the description of the culprit. In more recent literature there seems to be agreement that foils should be chosen based on their likeness to the description of the culprit, rather than matched to the suspect’s appearance (Brigham & Pfeifer, 1994; Tredoux & Chiroro, 2005; US Department of Justice, 1999; Wells, Seelau, Rydell, & Luus, 1994), to avoid the prototype face effect. This does however require a good enough description of the culprit in order to find suitable foils (Malpass, et al., 2007).

Not only are suitably similar foils necessary for a lineup, but there must also be enough foils to make a fair lineup, in order to reduce the incidence of a chance identification (Brigham & Pfeifer, 1994; Malpass & Devine, 1985; Malpass, et al., 2007). This will be discussed in the following section.


Lineup size

There is no global standard for lineup size. However, in South Africa, Paragraph 13(1) of the National Instructions 1 of 2007 state that a suspect must have at least seven foils (Naude, 2015). This number is somewhat arbitrary though (Wells, Seelau, Rydell, & Luus, 1994) which might be why there is no standard number. As discussed though, it is not enough to have seven foils, these foils need to be adequately similar to the suspect in order for the lineup to be fair. Thus lineup size is not simply how many members the lineup contains, but how many adequate members it contains (Ross et al., 2014; Tredoux, 1998).

Doob and Kirshenbaum (1973) were the first to research lineup size in relation to fairness. They researched a real lineup used in a recent case and gave descriptions of the suspect to mock witnesses who had to identify the suspect from the lineup. They were able to identify the suspect 11 out of 12 times (Malpass & Lindsay, 1999; Tredoux & Chiroro, 2005). They set up a measure for lineup fairness that used probabilities. If mock witnesses could identify the suspect more than an equal probability should dictate (for example if there were 5 lineup members they should each be selected 20% of the time), then the lineup was unfair. This method however does not consider the null foils, or foils who are so bad that nobody choses them (Tredoux, 1998). A lineup could have 10 members, of which 3 are null foils, and still be a fair lineup even if the suspect is chosen more than the chance proportion of 10%. It would only be unfair if the suspect is chosen more than 14% of the time. Bad foils should be considered when the size of the lineup is calculated (Malpass &Devine, 1985).

Further research has developed more sophisticated ways of calculating lineup size and fairness. These include functional size and effective size. Functional size represents the number of adequate, viable lineup members (Brigham & Pfeifer, 1994; Wells et al., 1979). It is calculated by dividing the total number of mock witnesses by the number of mock witness who chose the suspect (Brigham & Pfeifer, 1994; Tredoux, 1998). A higher number indicates a fairer lineup. This is a better calculation than the Doob and Kirshenbaum method because if there are null foils, they do not affect the functional size (Malpass & Lindsay, 1999; Wells et al., 1979).

Effective size adjusts its calculations for the inadequate foils (Malpass, et al., 2007; Ross et al., 2014; Tredoux, 1998). It has a distribution between 1 and k, the number of lineup members, and decreases towards 1 as lineup members are chosen more frequently than chance would dictate (Malpass, et al., 2007; Ross et al., 2014). For effective size, foils that are chosen below chance are not considered adequate foils (Brigham & Pfeifer, 1994). There is however no sampling distribution for effective size so statistical conclusions are difficult to make using this number (Malpass, et al., 2007; Ross et al., 2014) and there is no standard “good enough” effective size to measure lineups against. Despite this, effective size has been found to discriminate well between fair and unfair lineups (Brigham & Pfeifer, 1994).



Suggestibility in a lineup procedure is when the lineup administrator suggests to the witness the identity of the suspect, or pressures the witness to make a certain identification (Lindsay, 1994). Suggestion is the cause of more miscarriages of justice than anything else (Malpass & Devine, 1984), thus, whether conscious or unconscious, it is prohibited (Malpass & Devine, 1984; Tredoux & Chiroro, 2005). Unfortunately, merely holding a lineup could be suggestive because it implies that the police have a reason to believe that someone in the lineup committed the crime (Malpass & Devine, 1984).

Suggestion can be in lineup construction or procedure (Malpass & Devine, 1984). Suggestible lineup construction is when the lineup is so small or when the foils and suspect are not similar enough that it suggests the identity of the suspect (Malpass, et al., 2007; Tredoux, 2011). A case to illustrate this was in Minneapolis where the suspect was the only black member of an otherwise all-white lineup (Lindsay, 1994). This is clearly suggestive.

The following section will focus on procedural suggestion, since we have already discussed lineup size and foil similarity.

The investigating officer should not be present for the lineup and the administration officer should not know who the suspect is (Clarke, 2012; Davies & Griffiths, 2008; Naude, 2015; Tredoux & Chiroro, 2005), as confirmation bias and expectancy effects through non-verbal behaviour could inadvertently cue the witnesses’ identification (Clarke, 2012; Tredoux & Chiroro, 2005; Wells, Seelau, Rydell, & Luus, 1994).

Bias can also occur through biased instruction and verbal communication (Lindsay, 1994). Nothing should be said that suggests the identity of the suspect (US Department of Justice, 1999). Research shows that when the culprit is present, biased instructions lead to a 100% rate of making a selection, compared to 83% of participants from the unbiased instruction group (Malpass & Devine, 1984). Unbiased instructions also lead to no errors in this condition compared to 40% error under biased instruction (Malpass & Devine, 1984). When the culprit is not present, 78% made a selection under biased instructions, compared to only 33% under unbiased instructions (Malpass & Devine, 1984). This shows the effect of biased instruction on identification outcome.

An example of this is the case of the Eikenhoff three who were convicted of murder partly based on lineup identifications (Smalberger, J.A., Grosskopf, J.A., Mpati A.J.A., in Ndweni, Bholo, Gavin v S). Witnesses were shown photographs of the suspects before the lineup and told to identify them (Smalberger, J.A., Grosskopf, J.A., Mpati A.J.A., in Ndweni, Bholo, Gavin v S). It later transpired that they were not the true culprits and they were released (Smalberger, J.A., Grosskopf, J.A., Mpati A.J.A., in Ndweni, Bholo, Gavin v S). This was a biased lineup because witnesses should not see in any form or be described the suspects before the lineup (Davies & Griffiths, 2008; Naude, 2015), according to Paragraph 14(1) of the National Instruction 1 of 2007. This paragraph also states that witnesses should not see each other before the lineup so as not to influence each other (Naude, 2015; Tredoux & Chiroro, 2005). Officers must also instruct witnesses that the culprit may not be present in the lineup (Ainsworth 1998) as this causes fewer misidentifications, and lowers the rate of relative judgements, or choosing by process of elimination (Davies & Griffiths, 2008; Tredoux & Chiroro, 2005).

Suggestion impacts on all areas of lineup, and thus should be carefully controlled by following strict guidelines (Malpass & Devine, 1985).


Lineup fairness discussions are not objective and can be confusing (Malpass & Devine, 1985; Wells et al., 1979), especially for police officials who have not been trained sufficiently (Brigham & Pfeifer, 1994). This stressed the importance of adequate training to avoid unfair and biased lineups. It is beyond the scope of this essay to discuss the format of lineups in depth, but research indicates that sequential lineups are more fair than simultaneous lineup (Tredoux, 1998), because they yield better results and fewer misidentifications (Tredoux & Chiroro, 2005). One final point is that because of high rates of error in lineup identifications, even if all recommendations are followed, the evidence from lineup identifications needs to be tested rigorously in court (Tredoux & Chiroro, 2005).




Ainsworth, P. B. (1998). Psychology, law and eyewitness testimony. England: Wiley.

Brigham, J. C., & Pfeifer, J. E. (1994). Evaluating the fairness of lineups. In D. F. Ross, J. D. Read, & M. P. Toglia (Eds.), Adult eyewitness testimony: current trends and development (pp. 182-200). USA: Cambridge University Press.

Clark, S. E. (2012). Costs and benefits of eyewitness identification reform psychological science and public policy. Perspectives on Psychological Science7(3), 238-259.

Davies, G., & Griffiths, L. (2008). Eyewitness Identification and the English Courts: A Century of Trial and Error. Psychiatry, Psychology and Law15(3), 435-449.

Doob, A. N. and Kirshenbaum, H. M. (1973). Bias in police lineups partial remembering. Journal of Police Science and Administration, 1, 287-293.

Lindsay, R. C. L. (1994). Biased lineups: Where do they come from?. In D. F. Ross, J. D. Read, & M. P. Toglia (Eds.), Adult eyewitness testimony: current trends and development (pp. 201-222). USA: Cambridge University Press.

Malpass, R. S., & Lindsay, R. C. (1999). Measuring lineup fairness. Applied Cognitive Psychology13(1), S1-S7.

Malpass, R. S., Tredoux, C. G., & McQuiston-Surrett, D. (2007). Lineup construction and lineup fairness. In The handbook of eyewitness psychology, Vol II: Memory for people (pp. 155-178). New Jersey: Lawrence Erlbaum Mahwah.

Malpass, R., & Devine, P. G. (1984). Research on suggestion in lineups and photospreads. In G. L. Wells, & E. F. Loftus (Eds.), Eyewitness testimony: Psychological perspectives (pp. 64-91). USA: University of Cambridge Press.

Malpass, R., & Devine, P. G. (1985). Measuring the fairness of eyewitness identification lineups. In S. M. A Lloyd-Bostock, & B. R. Clifford (Eds.), Evaluating witness evidence: Recent psychological research and new perspectives (pp. 81-102). Bath, UK: Wiley.

Marwitz, D. B., & Wogalter, M. S. (1988, October). Bias in photospreads of faces: a comparison of two lineup construction methods. In Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 32(9), 541-543.

Naude, B. (2015). Ensuring procedurally fair identification parades in South Africa. South African Criminal Journal, 188-203.

Ndweni, Bholo, Gavin v S (1999) 4 All SA 377 (A)

Ross, S. J., Tredoux, C. G., & Malpass, R. S. (2014). Evaluating eyewitness testimony of adults. In The handbook of forensic psychology (pp. 513-559). New Jersey: Wiley.

Segomotso v S (2001) 1 BLR 125 CA

Smith, S. M., Lindsay, R. C. L., & Pryke, S. (2000). Postdictors of eyewitness errors: Can false identifications be diagnosed?. Journal of Applied Psychology85(4), 542.

Tredoux, C. G. (1998). Statistical inference on measures of lineup fairness. Law and Human Behavior22(2), 217.

Tredoux, C. G. (2011). Inaugural lecture: Pragmatic Psychology and the perils of eyewitness identification. Retrieved from University of Cape Town:

Tredoux, C. G., & Chiroro, P. M. (2005). Eyewitness testimony. In C. Tredoux, D. Foster, A. Allan, A. Cohen, & D. Wassenaar (Eds.) Psychology and law, (pp. 193-225). Cape Town: Juta.

U.S. Department of Justice. (1999). Eyewitness evidence: A guide of law enforcement. Washington, DC: Author.

Wells, G. L., Leippe, M. R., & Ostrom, T. M. (1979). Guidelines for empirically assessing the fairness of a lineup. Law and Human Behavior3(4), 285.

Wells, G. L., Seelau, E. P., Rydell, S. M., & Luus, C. A. E. (1994). Recommendations for properly conducted lineup identification tasks. In D. F. Ross, J. D. Read, & M. P. Toglia (Eds.), Adult eyewitness testimony: current trends and development (pp. 223-244). USA: Cambridge University Press.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s