Lineup Construction and Evaluation
Put together a lineup. Read the eyewitness guide for information on how to put together a lineup.
Go here: http://www.dc.state.fl.us/AppCommon/
Choose a guy (search on some characteristics, or a name).
Then find fillers to match
Copy and paste into a word document
Print.
Show to friends, collect data according to instructions provided in the below link.
Read: http://eyewitness.utep.edu/consult05B.html
Calculate lineup bias: http://eyewitness.utep.edu/documents/bias-calc.xls
Write about your experience and findings here.
This was a very interesting and enlightening exercise. First, creating the actual line-up was very difficult for me. I first decided to choose subjects of my own race to eliminate by own race bias, thus being able to choose subjects that looked similar. Secondly I went through several pages of photos selecting candidates. There were several instances in which I selected a subject for the line-up but once I pasted the photo into the array I was immediately aware of what I perceived as stark dissimilarities between the photos. I took me quite a few attempts before I was able to develop a six person photo array that I felt was appropriate to show to others.
I then conducted three different tests with five participants. In the first test I told each participant some very basic descriptive information of the suspect (ex. Caucasian, age 25-35, dark eyes, shaved head). I did not tell them any time of background information, such as what the suspect was wanted for. Only one out of the five participants was able to correctly identify the subject and the other four all guessed differently.
For the second test I interviewed the same five people and told them that the suspect was wanted for a brutal double murder. No one correctly identified the actual suspect. Three of the five subjects choose the same person incorrectly, but were very certain in their choice. The other two each choose a different candidate.
For the final test I told the same five subjects that the suspect was wanted for possession of drugs and was of Hispanic decent. Four of the five subjects selected the same candidate incorrectly. The fifth participant choose the correct candidate after quite a bit of hesitation.
The line-up bias for each of the studies:
-Study 1:0.401878
Proportion identifying the suspect (p) .200
q = 1 - p .800
n 5
Standard Error of p .179
Chance expectation: .167
Critical Ratio for diff from Chance: 0.186
(.05 requires 1.96)
(.01 requires 2.58)
-Study 2: 0.401878
Proportion identifying the suspect (p) .000
q = 1 - p 1.000
n 5
Standard Error of p .000
Chance expectation: .167
Critical Ratio for diff from Chance: #DIV/0!
(.05 requires 1.96)
(.01 requires 2.58)
-Study 3: 0.401878
Proportion identifying the suspect (p) .200
q = 1 - p .800
n 5
Standard Error of p .179
Chance expectation: .167
Critical Ratio for diff from Chance: 0.186
(.05 requires 1.96)
(.01 requires 2.58)
This activity was quite alarming to me because of the obvious inaccuracies of line-ups. Based on the results of my study I would be very hesitant to accept of line-up as an investigation tool. From my own research it appears that these types of inaccuracies are not uncommon in line-ups, however there are still a widely used investigative tool in the criminal justice system.
After looking at the unfair and poorly put together lineups in class, I wanted to make sure I did a good example of how a lineup should look. I chose a middle aged Caucasian male with short brown hair around the height of 6 foot. I did this because at first it would be difficult choosing considering that a white guy with brown hair is not that uncommon, I wanted the mock witnesses to not be stumped, but to have to look deeper into my descriptions. My results are as follows:
Number of mock witnesses: 10
Number of mock witnesses that chose correct: 8
Number in lineup: 6
Exact probability of these results 0.000019
Proportion identifying the suspect (p) .800
q = 1 - p .200
n 10
Standard Error of p .126
Chance expectation: .167
Critical Ratio for diff from Chance: 5.007
(.05 requires 1.96)
(.01 requires 2.58)
I was not too surprised by my results because I thought I did a fairly good job in describing the suspect. I think this is a job that, in reality, is very difficult. You want the witness to choose the correct guilty suspect, but you also do not want to influence the choice of an innocent person by putting different looking fillers in or making the decision too hard.
I found this activity to be very difficult because I had no idea how difficult it would be to find 6 men who are 5'10'' to 6' with brown hair. After sitting through class today, it made me realize that setting up lineups are more difficult than I previously thought. I had to find men with the same general head shape but not too close to the "suspect". The lecture in class today and this activity give me more respect for those who make lineups for a living and cops in general.
1. Enter the total number of mock-witness choices here: 12
2. Enter the number of mock witnesses who chose the suspect here: 7
3. Enter the total number of persons in the lineup here: 6
Note: Fields in which data can be entered are shown in boxes.
Other fields are calculated fields.
Exact probability of these results 0.001137
Proportion identifying the suspect (p) .583
q = 1 - p .417
n 12
Standard Error of p .142
Chance expectation: .167
Critical Ratio for diff from Chance: 2.928
(.05 requires 1.96)
(.01 requires 2.58)
I wasn't really surprised by my findings because I thought that I set up the lineup in a way that could be solved. I also described the suspect as close in detail as I could without completely giving it away. I really wonder how this activity would fare if the details of the suspect were more broad and whether or not the results would be the same.
I thought that this assignment was fun and difficult. Its not everyday that you get to put together a line up and have your friends figure out who it was. It was hard to find people that looked the same. There were people who had similar traits to that of my "suspect', but were far enough apart that it would be to easy. It showed that when putting together you really need to know your craft. Its not just throwing pictures together. The person putting these pictures needs to make sure that the suspect can only be identified by the witness and not random people off of the street.
My suspect was 6'2 white male with shorter dark hair and a rounded face. This was my bias calculation:
Exact probability of these results 0.054266
Proportion identifying the suspect (p) .400
q = 1 - p .600
n 10
Standard Error of p .155
Chance expectation: .167
Critical Ratio for diff from Chance: 1.506
(.05 requires 1.96)
(.01 requires 2.58)
I am not that surprised that the numbers came out as they did. I tried to get all of my suspects to look pretty similar and not have the bad guy stick out. Therefore when people had their different answers and comments, it was not that surprising to me. We could describe the suspect as we wanted which meant we could make it broad or very specific and that may have resulted in us making it to easy or very hard for our mock witnesses.
I found this assignment somewhat difficult to accomplish at first. Listening in class today I found that creating a lineup is very difficult for investigators to create to ensure that the suspect is not to easily identified and that the others are enough alike to create a good workable lineup. I was looking for individuals who are white six to six feet tall, short brown hair with large rounded eyes. To find my matches it took me awhile to find individuals who I thought would fit my lineup and help me make a workable lineup.
Exact probability of these results 0.001137
Proportion identifying the suspect (p) .583
q = 1 - p .417
n 12
Standard Error of p .142
Chance expectation: .167
Critical Ratio for diff from Chance: 2.928
(.05 requires 1.96)
I'm not surprised that my calculations came out the way they did. My friends told me that after hearing the description it was pretty easy to find the person in the lineup.This assignment proved to me that lineups are by no means assignment in and investigation, and why lineups can be very flawed as evidence in criminal case
I thought this experience was confusing but very enlightening once I figured it out. I had not thought about it much in the past, but there are so many people that can fit generic descriptions. This experience and activity has only confirmed and supported my already still beliefs about lineups and eyewtiness memories/stories/beliefs. Overall, I have learned from this activity that puting together a good lineup is difficult no matter what the desciption of the person is. As fot he eyewtiness, I can not imagine the difficulty they have picking out a person in a line up when they all look the same and are often wearing the same thing. It has raised even more questions to me such as how many people choose the suspect striclty because of luck and not because of their actualy account with the person.
My suspect was described as 6 ft. tall, hispanic male with short dark hair and presumably brown eyes. My results were as follows:
number of mock witness choices: 12
number of mock witnesses that chose the suspect: 7
total number of persons in lineup: 6
exact probablity of these results: .001137
Proportion identifying the suspect (p) .583
q = 1 - p = .417
n = 12
standard error of p = .142
chance expectation = .167
critical ratio for diff. from chance = 2.928
The results I acquired for this activity were kind of surprsing. I the task of creating a decent lineup was failry daunting, but apparently it was not that bad of a lineup. I truly think picking out a good lineup, participating as an eyewitness in a lineup situation, or being in a lineup has a lot to do with "chance". I simply do not think lineup identifiaction should solely incarcerate someone.
For my line up I choose six pictures from the Florida database. I looked for the following criteria:
Blue eye color, Height 5,9ft. to 6,5ft., Age range 22 to 25 years, white, male, blonde.
I choose 6 pictures of men of whom I thought it would be hard to identify just based on this description.
After reading the website about the mock witness procedure, I tried to follow the steps listed on the website. I selected the population of participants. Afterward I provided them with the information that I have chosen to give to the participants. I did not keep in mind while doing the procedure with the participants that it is important which information is given to the mock witnesses. Especially after hearing more about this topic today in class I came to the conclusion that I might have given away a too detailed description of the “suspect”, which might have made it easier for some people to identify my suspect.
I still tried to work with my collected data and it was very interesting to experience and learn from the mistakes I had made. One of the things I learned from this assignment is, that I would have to be more careful about the line up pictures I pick if I had to compose a line up again. My suspects might look similar but they are still too diverse to proof the real eyewitness's memory.
Another mistake I made when I chose the pictures was that I mixed inmate photos and booking photos.
Also, I did not pay attention to the babyfacedness criteria when putting my line up together on the weekend. If I would do another line up, I would pay a lot more attention to these little things which are important for the mock witness procedure, so that the pictures would be equally picked among the participants. In my experiment, all of the pictures got picked, but not equally distributed. My suspect got picked more often than the other photos.
The total number of mock-witnesses was 13. Mock-witnesses who chose the right suspect were 5.
Probability = 0,038492 p= ,385 q= ,615 SE = ,135 chance expectation = ,167 critical ratio = 1,615
To begin, I am going to admit my fatal mistake. When searching the registry of criminal faces I unwittingly decided to pick faces all connected with embezzlement. This was a massive error because not too many white male faces with red hair and brown eyes appear, but I guess I just wasn't thinking clearly enough to consider that not choosing a crime to link the pictures together would provide more pictures. But it doesn't matter now. I created my lineup and showed it to people before I realized my mistake, so we must roll with the results I have, and they are...
My lineup sucks.
Because most of my friends at UNI were incredibly busy today, I posted my lineup on Facebook (http://www.facebook.com/notes/brionna-sommers/experiment-for-my-psych-and-law-class-please-participate/10150333123924233?ref=notif¬if_t=note_comment). I gave directions on what to do and provided a "witness statment" (my vague description of my suspect). Then, I added six pictures of possible suspects and waited for comments. Admittedly, there is room for influence because mock witnesses could look at previous guesses, but not until the bottom of the page, so it is unlikely that many of them were influenced by each other.
Nine people commented on my lineup experiment. Out of that nine, six of them guessed the suspect, two picked a filler who looks similar, and one picked some random guy for reasons I don't quite understand. I popped my statistics into the spreadsheet provided above to determine my lineup bias. According to the bias-calculating spreadsheet, "If the proportion of mock witnesses identifying the suspect reveals bias, beyond a chance probability of .01, the critical ratio will exceed 2.58." My critical ratio: 3.182. Needless to say, my lineup is incredibly biased.
I thought that this was a fun exercise because it gave us a chance to actually put together our own line up and see how easy or hard it would be to actually put together a line up. I typed in a last name and a height and weight ranging from 5’8 – 5’11 and a weight of an upper limit of 175 pounds. The first picture I came clicked on was a young black male with closely cropped hair. The next thing I did was search for pictures with the same characteristics in the search box to make sure there wouldn’t be an overly large amount of bias. It actually didn’t take that long to find 5 other pictures that were similar to the one that I had originally chosen. I was surprised to find at the amount of people in the database that had at least somewhat close characteristics to the picture I had picked. The only real difference that I tried to make was skin color but tried to have pictures with darker skin color still match other characteristics
After finding all of the pictures and putting them in order I showed them to 10 of my friends and gave them the characteristics of 5’8 – 5’11 a weight of 160-175 pounds, medium dark skin and ears that stuck out. There were 3 pictures that were medium dark skin and 3 that would have been considered lighter skin. Of the 10 people that I asked 3 of them correctly identified the correct person which was the 5th of 6 pictures in the lineup. This and one other picture were the most identified of the 6.
Proportion that identified the subject- .300
Q = 1-p - .700
N= 10
Standard error of P - .145
Chance expectation - .167
Critical Ratio for diff from chance- .920
.05 requires 1.96
.01 requires 2.58
I don’t feel like using lineups is a very good tool for investigations. This is because I feel like you have to be close to the crime to get a really good look at the offender and if you are that close you are most likely scared and not really paying full attention to what the offender looks like, therefore it would be hard to identify that person.
For this exercise, I wanted to give my audience a good description of my suspect. However, I knew my description was going to be somewhat broad and fitting for many of the average Americans out there. Therefore, there would be many specifics, however many people would fit this characteristic. I chose a White Male, 5'10"-6'2", 22-30 years old, Black hair, 225-240 lbs. Here are my results:
Exact probability: .000568
p=.636
q=1-p .364
n=11
SE p= .145
Chance Expectation=.167
CR for diff from Chance= 3.238
5%=.352-.921
1%=.262-1.011
After reading and learning about police lineups, I have learned how difficult an accurate police lineup can be. Fairness is a huge deal so the lineup itself is valid. Police are able to "screw up" lineups so easily, it makes it a difficult task to do it correctly. Many things have to be done on the lineup side of things as well as the identification and witness aspect in order for it to work as planned.
For my suspect I chose a 33yr old white male, 6'00", 200lbs, pale blue eyes that are sunk in, brown clean cut hair, smooth long face,and a slightly crooked nose. I chose him because he has an innocent looking face, a slight smirk, and no viable tattoos. He seems like a guy that could be blended nicely with other white men and not stand out too much.
mock witnesses: 10
mock witnesses that chose correct: 9
lineup: 6
Exact probability of these results 0.000018
Proportion identifying the suspect (p) .700
q = 1 - p .200
n 10
Standard Error of p .129
Chance expectation: .164
Critical Ratio for diff from Chance: 3.007
I think my list of suspects was chosen pretty good. None of my suspects directly stood out, yet you could still tell all of them apart. I gave a pretty detailed description of my suspect and I think it was helpful that I chose someone within my own race. However, many cases do not end this way and we often have trouble identifying different facial characteristics outside of our own race.
This assignment was both fun and interesting for me to put together. I thought creating the actual lineup was more difficult than I originally thought it would be. I wasn’t sure if I wanted to put together a lineup based on members of my own race or another. I decided to conduct a lineup of African Americans because I wanted to see if the cross-race effect could happen even in mock witnesses. I described the suspect as an African American male with black hair and brown eyes. The suspect is 23 to 30 years old, about 185 to 215 and stands about 5’9” to 6’1” tall. I said the suspect also had short cropped hair and a scar by one of his eyebrows (it took me a while to find ones that did). I thought these were pretty specific but I don’t think they made my suspect stand out too much either. My results were this:
Exact probability of these results: 0.005346
Proportion identifying the suspect (p)-.467
q = 1 - p-.533
n=15
Standard Error of p- .129
Chance expectation: .167
Critical Ratio for diff from Chance: 2.329
(.05 requires 1.96)
(.01 requires 2.58)
I was actually surprised by these results. I worked really hard to make my lineup unbiased but according to the bias calculation it was not good. I tried my best to make sure one suspect did not stand out from another but my mock witnesses said that it was somewhat easy to tell which one I was describing. This frustrated me because I really thought I did all that I could to make sure that my lineup wasn’t that easy. This exercise proved to me that lineups can be really unreliable and very biased if someone who has never seen the suspect can still pick them out of a lineup based solely on a verbal description.
Putting this lineup together was actually very time consuming because I was so intrigued with this exercise and found it to be rather fun. I finally decided on a male, white, brown hair (although bald in the picture) and blue eyes. It was actually somewhat difficult to go through all of these pictures and find someone that matched this criteria. I think the baldness is what made it difficult.
When explaining my criteria, I did not give more detail as to anything other than what this person looked like, and what I was looking for. Besides all of the previous facial features, I described him as about 5'11, 160 pounds, and was about 45-50 years old. I said that he had a very slender face with high cheekbones. I figured that this was enough information to get someones mind racing. I then decided to ask 12 participants (mock witnesses) if they could pick the suspect out of 6 photos. 5 of which are obviously fillers.
Number of participants (mock witnesses) = 12
Number of mock witnesses that guessed suspect number 1 (the right suspect) = 5
Exact probability = 0.028425
Proportion identifying the suspect =.417
Chance expectation = .167
Critical ratio = 1.757
q=1-p
1-.417=.583(q)
Standard error of p= .142
After doing this I realized that my only fault was sticking in a man that had a babyface with 5 other men that looked like cold blood killers. Baby face was looked a little younger than 45-50, but could have passed for a well aged man nonetheless.
I really enjoyed this assignment. It was somewhat difficult to find individuals that looked similar but yet were different enough that the six individuals were not carbon copies of each other. After a fairly good while of searching I found six males that I thought fit the bill. I began my search by typing in the last name of one of my good friends, alexander, and then finding 5 other people that have characteristics similar to the person I selected. My suspect was 5'8" with brown hair, brown eyes, close cropped hair, and 160 pounds. In my search for the fillers, I looked for guys that were 5'7" to 5'9", 160 - 180 pounds, white, male, brown hair, and brown eyes. A few unique characteristics that he possessed were a pale complexion, a strong chin, and a very symmetrical face. The identifiers that I game to the 12 friends that I asked to identify my suspect were as follows, pale skin, brown hair cropped close to the scalp, symmetrical face, and a strong chin. Of the 12 people that I interviewed, only two correctly identified the suspect. The other 10 individuals were split among two others in my lineup. The following are the stats computed by the spreadsheet you provided:
Exact probability of these results 0.296094
Proportion identifying the suspect (p) .167
q = 1 - p .833
n = 12
Standard Error of p .108
Chance expectation:.167
Critical Ratio for diff from Chance: 0.000
(.05 requires 1.96)
(.01 requires 2.58)
Confidence intervals
5% confidence interval equals -0.044 to 0.378
1% confidence interval equals -0.111 to 0.444
After doing this assignment, I'm not sure how reliable of a tool a lineup actually is. It is so easy to fudge the results in either direction. It would be easy to compile a lineup of people so similar that it would be impossible to pick out the right guy and at the same time it would be easy to create a lineup that causes the suspect to stick out like a sore thumb. The reliability of a lineup therefore is solely dependent upon the competency and ethics of the person creating it. After studying the material you have given us over the last few weeks I fear that I am going to be forever skeptical of eyewitness testimony.
In my line up I put 6 men with the description: male, white, black hair, 5 foot 9 inches, brown eyes 20 years of age, shaved hair, 170lbs, rounded face with bushy eye brows. I asked each of the 20 people to get an idea of the man after reading the description and then after they had a good idea I showed them the line up. Each person studied the lineup for awhile and covered up some guys to concentrate on others. My statistics looked like as fallowed:
Number of mock witnesses=20
Number of mock witnesses that guessed the right suspect=5
Exact probability= 0.129410
Proportion identifying the suspect =.250
Chance expectation = .167
Critical ratio = .861
q=1-p
1-.250=.750(q)
Standard error of p= .097
My lineup worked out fairly well and not many people got it. I did better understand how it is difficult to make a solid lineup as well for a victim to identify a perpetrator off of a lineup. It was a fun project and I got to see how different people reasoned about what guy they thought it was.
Doing this line up was very difficult. It took a long time to generate people I would feel comfortable for comparison without one sticking out. When doing the lineup I was actually shocked at what happened. I had asked people from multiple racial backgrounds and those who were black seemed to be more confident in their choices while those who were of other racial groups seemed to have a hard time telling them apart at all. This I am attributing to the fact that my line up was of black males.
The probability was .229874 and was within the 1 percent confidence interval. I take from these findings I made a sufficient lineup that would be appropriate to show if there was a real crime or suspect. I found it difficult as mentioned above; however, it would have been even more difficult to create an unbiased lineup with a group outside of my own race. Thus, I think that police departments need to be more aware of this biased that can occur, this is something that could be prevented by having officers of that nationality help create the lineup. If possible have mock witnesses from that race or the same race as the victim.
It is even more difficult for me to understand that fact that people go to prison purely on what someone saw. This showed just how easy it would be to make a line up biased, and this is done without much knowledge that this was occurring. If police and prosecutors are willing to use eye witness testimony as evidence everything should be done to guarantee that the witness is not contaminated. This means that everyone who has contact with witnesses need to understand and follow procedures to ensure that the evidence is not contaminated.
I enjoyed putting together a lineup for two reasons, first it was just fun to experience trying it, and second it was interesting to try to put together a line up for a guy who looked like a mix between father time and santa claus.
The only thing that I could really find difficult in this lineup were the pictures. It's hard to find people that don't have a dead giveaway background that they aren't in jail that look somewhat similar to this guy.
I chose a white male, brown hair, 6'00", 324 Pounds, 65-70 years old. I chose 10 participants if they could pick the suspect out of 6 people in a lineup.
# of participants: 10
People that guessed right: 3
Exact Probability: 0.155045
Proportion identifying the suspect: 0.300
Chance Expectation: 0.167
Critical Ratio: 0.920
q = 1-p
1-0.300=0.700
SE(p)=0.145
After I finished asking 10 people if they could identify the man out of the lineup it was too hard for them with his characteristics.
1. Mock witnesses = 17
2. Mock witnesses chose the suspect = 4
3. Total persons in lineup = 6
4. Exact probability of these results = .171639
5. Proportion identifying suspect =.235
q=1-p = .765 and n = 17
6. Standard Error of p = .103
7. Chance expectation = .167
8. Critical ratio for diff from chance = .667
9. Confidence intervals
5% = .034 to .437
1% = -.030 to .501
My lineup appeared to be a good one, as my chance expectation fell within both the 5% and 1% confidence intervals. Since it did not fall outside the given ranges, I have a lower lineup bias.
I found this activity to be quite fun! I first found a picture in the Florida correctional data base, of a person, that would be my suspect. Looking at the picture for a few minutes, I took it away and wrote down what my description of him would be. I then typed those variables into the database to get a list of other offenders that would meet the general criteria for white, male, 21-31 years old, 130-155 lbs. Going through the pictures I chose men that met the criteria, but also had dark, short hair and a receding widows’ peak, which were descriptions I had written down before searching for my fillers. Many of my fillers met this criteria and I heard many people sigh with discouragement when they flipped the paper over to examine the lineup. Two fillers and the suspect where selected more so than three other fillers. I found this interesting as one of the fillers that was selected by only one mock witness looks the most similar to the suspect and has a criminality about him. The criminality this filler possessed was that his chin was tilted upwards, which was also done by the suspect. The two fillers that were selected most often had very pronounced receding widow peaks, which may have stood out from the description more than the other data.
The idea of cops and detectives having to regularly do these mock lineups, with people, I think would be a great idea and practice. However, the high possibility of bias and error still shine a somewhat negative light on this procedure to me.
Asking people to participate was fun, and they rather enjoyed participating; something they probably wouldn’t enjoy had it involved a real case with a real suspect.
When I looked at the Lineup assignment I first thought this would be fun and pretty interesting. Being able to put my own line up together and then have other people try to find my suspect I felt would be pretty difficult for the others and myself. My suspect was a white male, short brown hair, 6'0, in his 30's and 200 lbs. It was kind of hard for me to really make it so one man didn't stick out more than the others. I even went through multiple photos to find the line up of just 6 people.
Number of mock-witnesses- 13
Number of mock witnesses who chose suspect- 8
Total number of persons in lineup-6
Exact probability of these result- .000308
(Proportion identifying the suspect) .6153846154
(q=1-p) .3846153846
(n) 13
Standard error of p .134932003
Chance expectation .167
Critical Ratio for diff from Chance 3.326
(.05 requires 1.96)
(.01 requires 2.58)
Confidence intervals
(5%) .351 to .880
(1%) .267 to .964
I feel my lineup was pretty well put together. No one in the lineup stuck out form one another and it wasn't all the way easy or hard to pick out my suspect. I really enjoyed doing this activity. It made me learn information about lineups but also I got to interact and do my own. By far my favorite assignment yet.
I did enjoy doing this assignment. Before class today, I would have put together a terrible lineup but after what I learned today, I understood that the lineup has to be put together the correct way so that not just anyone can identify the person. Putting together the lineup was not the easiest thing but it was quite fun and interesting and also time consuiming. I first started off by putting in my last name into the database and started searching. Then I started to put in a bunch of random information in which I used a black male. After making the lineup, I gave it to some friends to see if they can identify the criminal. I identified the person as 5'10" - 6'0", 180 - 200 lbs, black male with short hair and 20 - 25 years old. Giving this information, I figured that these details would not give any of my subjects a good chance to guess the criminal. After calculating the lineup bias, I found these results:
Exact probability of these results: 0.054266
Proportion identifying the suspect (p): .400
q = 1 - p: .600
n: 10
Standard Error of p: .155
Chance expectation: .167
Critical Ratio for diff from Chance: 1.506.
It was kind of hard to identify my individual. I think that my lineup was decent but I did realize that lineups would not be a very good idea if it was the only evidence that someone had unless they were 100% sure that they knew it was a certain person and all of the mock witnesses chose somebody completely different. That would be the only way in my opinion that you could use a lineup for evidence.
One thing that you must know about me is that I am horrible at math. I had no problems finding criminals for my lineup and people to help me out with the guessing. Then it came down to math. I think I did the portion identifying right. The number I got for that was .4. I had 6 suspects in my line up and had 5 people guess to see which one they thought it was. Two of the 5 answered correctly which I was surprised. I thought I had good descriptions but since only two people answered correctly I guess I was a little ahead of myself. I would type in the formula and everything but I must it up and my numbers would not match anything.
For line up I chose to have a caucasian male, 5'9, late 50's early 60's, and bald. I figured it would be easier to do than having some suspects with brown hair because it's so common.
While talking in class today about how some police officers are biased when it comes to putting together a lineup, I had to think and make sure that these suspects all looked the same. One of my guy friends had asked me why none of the suspects were african american and I told him it would be easier to identify people of the same race. In some ways it was difficult to put together the lineup because I would pick people with glasses but the suspect I chose to be the "criminal" didn't wear glasses so I had to change my line up a couple times to make sure that they all had the same characteristics as my criminal.
I thought this assignment was very fun, but yet much more challenging than I had originally thought. It was very interesting and challenging to put together my own lineup out of my suspect of choice and 5 other fillers that were similar to my suspect. The suspect I chose was a white male, 5' 08" tall, with a medium build. I stated all of this in my description of the suspect on my lineup sheet, as well as the fact that he had close cut, light colored hair, ears that stuck out a little bit, and a thick neck. These were all personal observations I made of my suspect. So, I had to find fillers with this similar makeup to put in my lineup. Once finished, I took my lineup sheet with me to work and had my 12 co-workers and bosses see if they could seriously try and determine who the suspect was. Going into this, I thought I had a rather unbiased lineup, I spent forever finding what I considered good fillers, yet it seemed as though they were all choosing a particular filler. It turned out that one filler had slight bruising on his face that I didn't notice originally when choosing him for my lineup. A few of my co-workers also chose another man I had in the lineup, seemingly just because he had some tattoos on his neck and face, even though I did not have that in the description of my suspect.
My results:
I'm sorry to inform that I do not have the Excel program on my home computer, so I wasn't able to open this lineup bias calculator. I even googled it, but I couldn't understand or retrieve enough solid equations of the formula to do it myself, so I will just list my results. I will try and go to the library tomorrow and post my results from this equation. I apologize and hope I do not get docked for it, since I did spend a great deal of time on this blog. My results were out of 12 co-workers/bosses, the number of people that chose the fillers and suspect are listed:
Filler (#1)- 0
Filler (#2)-3
Suspect (#3)- 1
Filler (#4)- 6
Filler (#5)- 0
Filler (#6)-0
The results clearly indicate that my lineup was very misleading towards an innocent filler! Honestly, I was pretty shocked when I saw the results weighed so heavy onto a single suspect, just for slight bruising that even I didn't catch! Obviously, my co-workers spent their time extrenuously evaluating these images. However, there could be another reason for this bias. Although I told my co-workers and bosses that it was important for them to decide alone and for themselves who they thought the suspect was, it was really hard in my workplace to not overhear or gossip about what other people were choosing. This may have lead to a bias of its own. I really enjoyed doing this activity and I learned how hard it can be to not only put a lineup together, but to prevent contamination of witnesses as well. Again, I apologize for not having the results calculated properly. I will try and submit them from a different computer tomorrow. This is the only one I have access to now, and I did not know I would need Excel to open the program. Thanks!
Lindsey Fails
I really enjoyed doing this mock lineup. I was fun to see if I would ever be good at making a lineup myself. When we talked about how lineups can be kind of a “make it or break it” situation for a lot of defendants it really got to me. I think it’s awful that some trials can have the one lineup they did as there only evidence, and I wonder what would happen if they made the rules a little more strict about evidence presented in court. Even when the judge tells the jury to disregard certain evidence sometimes, you can’t actually believe that they can just erase what they just heard or saw.
I was lucky in finding all my pictures within a half hour, and I didn’t realize how similar a lot of inmates look to each other. My description was a white male about 45-50yrs. old, about 190lbs, and had a wide forehead with semi-thick eyebrows. They all also had the same hair line. Just from looking at my pictures I thought I did a pretty good job before I had tested it out. My results were:
-number of people in mock lineup: 11
-number of people who guessed the perp: 3
-number of people in the lineup: 6
-probability of these results happening: 0.177656
-standard of error: 0.134
-chance expectation: 0.167
-critical ratio for diff from chance: 0.790
Out of my 11 participants, only 3 chose my designated perp. I had 5 people chose the person who I had originally thought people would pick anyway. Even though they all fit the description very well, I think that the one the most people chose had just a slightly more “criminal” look to him. It really showed how these negative stereotypes can affect witnesses and jury members.
I enjoyed this assignment. For my lineup, I looked up people in the database with my last name and chose a man from there. The man was white with brown hair, brown eyes, 5' 6", 150 lbs with round face, pointy ears, and was wearing round glasses. I chose 5 other men who could possibly fit the description, all of whom were wearing glasses, although some didn't have the round face or pointy ears. I asked 10 people to see if they could identify the suspect based on the information I gave them about his physical appearance, and out of the 10 people, 8 chose the correct person. I think this was because my description of his face was "round" and the other people in the lineup had more square jaws.
Number of mock witnesses: 10
Number of mock witnesses that chose correct: 8
Number in lineup: 6
Exact probability of these results 0.000019
Proportion identifying the suspect (p) .800
q = 1 - p .200
n 10
Standard Error of p .126
Chance expectation: .167
Critical Ratio for diff from Chance: 5.007
(.05 requires 1.96)
(.01 requires 2.58)
I really liked this assignment. Looking through the mug shots kind of creeped me out, however. The men all just look mean. It was kind of depressing.
Finding a person who you want to use as a suspect was an interesting experience. I just picked someone at random, and then created the description based on them. I picked white males that are grey or balding, with dark colored eyes, between 5'11'' and 6'2'' that was around 200lbs and in was 40-55.
I was surprised just how many men in the florida corrections system actually match that description. It was also hard to quantify bald or grey, since most people aren't completely bald or completely grey haired. There was also quite a range in how old the suspects appeared to be when I was looking through the pictures. Some of the 40 year olds looked younger, and some of the older men looked atleast sixty years old.
Instead of printing out my lineup and passing it around to people, I put my spreadsheet on facebook. This allowed me to reach more people and people outside of Cedar Falls. I told them to just message me their answers or text me with them. I got sixteen people to do this lineup for me while I was working and not actively seeking people out. This technique worked very well for me.
My critical ratio was -0.504. The thing that surprised me was that most people thought I had picked someone other than my real suspect. I chose suspect 2 to be the real criminal. Suspects 2, 5, and 6 each had two votes to be the criminal. Suspects 1, 3, and 4 had three votes each.
One thing that really surprised me was the certainty people had that their choice was the right one. They would say things like, "It is definately number six." and then would ask me if they got it right. When I told them I had actually picked number two to be my suspect, the people who didn't chose him were shocked. Each person thought their choice was the obvious answer.
This was a really fun activity, and it sparked a lot of great conversations. I would love to do this again to test more of my friends!
I felt that this exercise was a pretty good one. Being able to access Florida's database of the people who have been in jail was interesting to me.
What I went about doing for finding a suspect I wanted to pick was just search for white, between 5'9 and 6'3, brown hair, and blue eyes. I picked a guy who fit that description but had really short brown hair. He also had a mole on his neck which almost made me not want to pick him because then I would have to find 5 other men with same looks and a mole on their neck on top of it. But I did find 5 other men who fit the description which really surprised me.
I gave the friends I gave them this description: "White male, mid 20's, 180-200 pounds, short brown hair, blue eyes,and a mole on his neck". I then ask them if they could identify the suspect given the description. My suspect was number 3 out of the 6 in the lineup. I asked 13 of my friends. Number 1 filler was chosen 2 times. Filler number 2 was chosen 2 times. The suspect was chosen 2 times. Filler #3 was chosen times 3 times. Filler number number Four was chosen 1 time. Filler #5 was chosen 3 times. This goes to show that I either had a very good lineup in which only the witness would recognize the suspect or I made the lineup so hard that no one could recognize him if they knew him. But I think the first of the two is a better indicator of what happened.
After plugging in my results into the excel worksheet that calculates if there was a bias in the lineup, it said that my lineup had nowhere near a bias. It says that a chance probability of .05 or less is sufficient to indicate that there is bias in a lineup. My chance probability is .167, which far exceeds .05. These results show me that I did, in fact, make a pretty good lineup. If I did it again, I would probably try and find more friends to look at the lineup to see if maybe with more sample, the suspect could be identified easier or not.
This exercise was interesting to do because choosing people for the lineup ended up being tougher than I expected. I chose a white male with closely-shaved brown hair, brown eyes and about 165 pounds. I thought he would be easy to match with other white males with brown hair. Face, eye, nose, mouth and jaw shape are very different on each individual so it was hard to find 5 other men who looked similar enough to my suspect. Here are my results:
-10 mock witnesses
-3 chose the correct suspect
-Exact probability of these results: 0.155045
-Chance expectation: .167
-Critical Ratio for diff from Chance: 0.920
-Confidence intervals
5% confidence interval equals 0.016 to 0.584
1% confidence interval equals -0.074 to 0.674
I think I ended up doing a pretty decent job since only 3 people chose correctly. The chance expectation is within the 5% and 1% intervals so that shows that I had a lower lineup bias.
My mock witness’ responses were pretty well distributed between the others in the lineup. It was surprising to me that after reading the brief description and looking at the pictures once, the mock witnesses wanted to look at the description again, sometimes multiple times. One person had trouble deciding between 3 of the fillers, eventually deciding on the real suspect. They also were surprised when they found out they had chosen the wrong person and would try to justify their reasoning based on the description I gave. I think this exercises shows that lineups can be faulty and they can be biased because of how the description is worded or presented.
I really enjoyed this assignment, but like the others I too found it more difficult than I had initially thought it would be. I began my search using my last name, and came up empty. I then thought, “Hey, why not choose someone who looks like me!” (I obviously didn’t read the instructions carefully because after reading other posts and re-reading the directions, I read the portion about picking a guy. Hope it is okay that I picked a gal!) I selected white, female with blonde hair, and blue eyes for my criteria search and went from there. It was quite difficult to pick 6 females who had the same characteristics. I think I went copy, pasted and then deleted about 15 different individuals before I had my final line-up.
I work at Texas Roadhouse, and I knew that my fellow roadies would be more than willing to help me with this assignment. I individually asked 15 of my co-workers to read the description I had provided: A white, female with long blonde hair, blue eyes and an oval shaped face. She is between the ages of 20 and 30, about 5’4’’ to 5’7’’ tall and weighs between 140-160 pounds. Two of the fill-in females in the lineup were never chosen. One of the fill-in women in the lineup was chosen 8 times.
Calculation of lineup bias:
Number of mock witnesses: 15
Number of mock witnesses who answered correctly: 3
Total persons in lineup: 6
Exact probability of these results: 0.236256
Proportion identifying the suspect (p): .200
q=1-p: .800
n: 15
Standard error of p: .103
Chance expectation: .167
Critical ration for diff from chance: 0.323
I did a pretty decent job at putting together a non-biased lineup, turns out. After asking all my friends individually about who they thought the suspect was, everyone came together and huddled around the pictures and descriptions and discussed why they thought the way they did. I didn’t let people change their answers, because then they would have been biased from listening to others, however it was exciting listening to everyone debate on why they were right and others were wrong. At the end of the night when I told everyone what the answers were, there were groans, looks of shock and a couple high fives among the “winners”. Participating in this lineup experiment really opens your eyes. We are biased to ourselves in thinking that we are always correct (most of the time), but as we could see, 15 people were giving the SAME characteristics, however different people were chosen for different reasons from the lineup. We are all psychologically different in the way we think and process things and therefore, I don’t think lineups seem to be one of the best options to use in trial when trying to convict a potential criminal.
I would first like to say that I was very confused when it came to forming the lineup. There were so many different possibilities that narrowing it down was difficult. I chose a white male, 6’ to 6’4”, 20 to 28 years old, 160-200lbs, and brown hair. I chose these characteristics because they seem to be common and I wanted to see how people would decide on a suspect. Making the lineup itself was difficult for me because I wasn’t really sure who I wanted for fillers and who I wanted the suspect to be. By going through the many pages of possibilities, I finally formed a lineup that I thought was fair yet testing to any witnesses, meaning they would need to pay close attention. One error I found that I made was that my description was too broad and most of my witnesses were stumped. I then gave a little more of a description by saying that their hair was about 1 to 2 inches long and they had lighter colored eyes.
Mock Witnesses: 10
Number of Mock Witnesses that chose correct: 4
Number in Lineup: 6
Exact probability of these results: 0.054266
Proportion identifying the suspect (p):.400
q=1-p
n 10
Standard Error of p: .155
Chance expectation: .167
Critical ratio for diff from chance: 1.506
(.05 requires 1.96)
(.01 requires 2.58)
I don’t know if it was my description or how serious my mock witnesses were taking the lineup that caused my results. I felt that I had given enough of a description to help them narrow down the suspect without fully giving it away. My lineup seemed close yet the extra description I had provided, if they would have looked closely, they would have found the suspect. Then again, I’m not sure what I would do in that situation either. This whole exercise has taught me that lineups aren’t as easy as they are portrayed. I was not aware of all the different things that are required in order to make a lineup as accurate and fair as possible. From this, I also decided that it is probably not a good idea for me to ever make a lineup because it is way more difficult than most of us think.
I had a lot of fun doing this assignment! It was very interesting to see the reasoning behind the choices made by my co-workers who picked out their "murderer" out of the line-up. My suspect/description given to mock-witneses was a 35-45 year old caucasion (everyone I work with is white so I wanted to exclude racial bias)with dark, short hair and had pointy ears. I made this description after first choosing my suspect. Afterwards I found the fillers and tried to choose men who most closely matched the description. (there were a lot more matches than I thought there would be!)I had the chance to ask eleven people. After giving them the description, only 4 of them chose the right suspect. The interesting thing was that there was one filler who had kind of had distinctively shaped eyebrows that otherwise wouldn't have stood out at all and didn't even have as pointy of ears, but was chosen 6 times. The reason for this was "he looked mean". It is very interesting that the difference between how bushy or angled someones eyebrows might be could be what costs them jail-time! (note to self: keep eyebrows managed.)
Anyways, my results were:
Exact probability of these results: 0.071062
Proportion identifying the suspect (p): .364
q = 1 - p .636
n - 11
Standard Error of p .145
Chance expectation: .167
Critical Ratio for diff from Chance: 1.358 (.05 requires 1.96)
(.01 requires 2.58)
Confidence intervals
5% confidence interval = 0.079- 0.648
1% confidence interval= -0.011 to .738
It is obvious that lineups are NOT a good source of evidence, and bias is often difficult to avoid, especially when negative stereotypes have so much effect on judgement! --very interesting assignment! :)
This was a lot of fun to do. It was neat to search family names and see what type of people pop up. There were zero people with the last name Jondall in the system, but the Mills side of my family had a couple dozen people. I ended up choosing, for my lineup, just your average 40 year old male that is Caucasian and around six foot tall. He had somewhat of a mustache/beard, but it was kind of hard to see. I easily found five other guys that resembled him pretty well. He was my number six suspect on the sheet and was only chosen five out of the thirteen times. I feel that is good, because it wasn’t a dead ringer and wasn’t way too easy. Numbers two, three, and five were each picked three times. My perp was number six and he was picked five times. What I really learned is how scary it could be to potentially be picked out of a lineup as a suspect. You may not be the person who did the crime, but some random person may think you were the one that they saw. It really shows that the memory and mind can sometimes coerce you into believing something that may not even be true. These lineups are clearly not enough evidence to try someone with and I would hope the jury could see that.
My data showed:
Exact probability of these results 0.038492
Proportion identifying the suspect (p) 0.385
q = 1 - p 0.615
n 13
Standard Error of p 0.135
Chance expectation: 0.167
Critical Ratio for diff from Chance: 1.615
(.05 requires 1.96)
I really enjoyed this assignment as it helped proved some of the biases that are possible in lineups.
My suspect that I found was in the height limits of 5’4 to 5’8 with shortcut blonde hair, and blue eyes. Also a noticeable feature is a dimple on his chin. As I looked began looking through the database I thought it would be tough to find suitable unbiased lineup matches. But to my surprise I found a few suitable candidates 5 to be exact. And from there on I conducted what I have learned is called mock witnesses. It went alright as 15 people participated. I give them an accurate description and let them make a choice only 5 guessed it which could mean a few things as we learned today.
The results say
Exact probability: .062372
Standard of error : .122
Critical ratio: 1.369
This to me looks like a pretty decent lineup and not to biased as well as only a third of the witnesses identified them.
I expected this assignment to be easy, but I found I was mistaken. It ended up being a fairly difficult assignment. In class today, I learned that creating a lineup is a challenge, but I underestimated how much of a challenge it would be. The suspect I chose was a 31-year-old black male who had the same last name as me. The criteria that I used to create my lineup consisted of other black males who were in their late 20s to early 30s. Other criteria that I used to create my lineup, in order to kind of match the suspect, was height (5’9”-6’0”), weight (190-210 lbs.), hair color (black), and eye color (brown). So, once my options for fillers was narrowed down with these criterion, I then looked for similar facial features like large, pug nose, large lips, and bags under eyes. I did not realize how hard it would be to find fillers, though, who were similar to the suspect yet not too similar. I also did not realize how hard it would be to word the description. Overall, however, I think I created a pretty fair lineup.
My results were:
Total number of mock-witnesses: 10
Number of mock witnesses who chose the suspect: 2
Total number of persons in the lineup: 6
Exact probability of these results: 0.290710
Proportion identifying the suspect (p): .200,
q = 1 - p: .800, n: 10
Standard Error of p: .126
Chance expectation: .167
Critical Ratio for diff from chance: 0.264
Confidence intervals:
5% confidence interval equals: -0.048 to 0.448
1% confidence interval equals: -0.126 to 0.526
According to my results, I created a fair lineup since my critical ratio did not exceed 1.96 or 2.58 and my chance expectation was within the 5% confidence interval and within the 1% confidence interval. Most of my friends needed to look at the description a second time before choosing someone from the lineup. In addition, none of them were confident with their choices. I have learned that it is very hard to create a fair lineup, and that they are not a reliable method to use to identify a suspect.
This was one of the most difficult tasks that i have ever come across thus far. Well i should say challenging assignments, because putting together these line ups were so hard. Finding suspects, let alone fillers was quite the challenge because everyone of these women that ive chosen had very similar characteristics, height, and weight and especially hair. i have to admit that in the begining i was very scattered and lost when trying to put these suspects together and find fillers that also looked like the suspect without their pictures standing out any more then the other. here are my results....
Instructions.
1. Enter the total number of mock-witness choices here: 20
2. Enter the number of mock witnesses who chose the suspect here: 12
3. Enter the total number of persons in the lineup here: 6
Note: Fields in which data can be entered are shown in boxes.
Other fields are calculated fields.
Exact probability of these results 0.000013
Proportion identifying the suspect (p) .600
q = 1 - p .400
n 20
Standard Error of p .110
Chance expectation: .167
Critical Ratio for diff from Chance: 3.956
(.05 requires 1.96)
(.01 requires 2.58)
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .05, the critical ratio will exceed 1.96.
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .01, the critical ratio will exceed 2.58.
A chance probability of .05 or less is sufficient to indicate bias in the lineup.
A rule of thumb is that no more than 50 mock witnesses should be used.
Confidence intervals
5% confidence interval equals 0.385 to 0.815
1% confidence interval equals 0.317 to 0.883
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 5% confidence interval, we can be 95% certain that the
lineup is biased.
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 1% confidence interval, we can be 99% certain that the
lineup is biased.
This assignment was quite the challenge. when i first started going through this assignment it felt like it was going to be as difficult as it seemed. Going through those line ups were pretty critical because your trying to find photos of people who favor other characteristics as others, at the same time know identifying differences in the photos that would stand out that wouldnt in other photo's.
These were my results:
Exact probability of these results 0.000013
Proportion identifying the suspect (p) .600
q = 1 - p .400
n 20
Standard Error of p .110
Chance expectation: .167
Critical Ratio for diff from Chance: 3.956
(.05 requires 1.96)
(.01 requires 2.58)
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .05, the critical ratio will exceed 1.96.
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .01, the critical ratio will exceed 2.58.
A chance probability of .05 or less is sufficient to indicate bias in the lineup.
A rule of thumb is that no more than 50 mock witnesses should be used.
Confidence intervals
5% confidence interval equals 0.385 to 0.815
1% confidence interval equals 0.317 to 0.883
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 5% confidence interval, we can be 95% certain that the
lineup is biased.
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 1% confidence interval, we can be 99% certain that the
lineup is biased.
This assignment was quite the challenge. when i first started going through this assignment it felt like it was going to be as difficult as it seemed. Going through those line ups were pretty critical because your trying to find photos of people who favor other characteristics as others, at the same time know identifying differences in the photos that would stand out that wouldn’t in other photo's.
Instructions.
1. Enter the total number of mock-witness choices here: 20
2. Enter the number of mock witnesses who chose the suspect here: 12
3. Enter the total number of persons in the lineup here: 6
Note: Fields in which data can be entered are shown in boxes.
Other fields are calculated fields.
Exact probability of these results 0.000013
Proportion identifying the suspect (p) .600
q = 1 - p .400
n 20
Standard Error of p .110
Chance expectation: .167
Critical Ratio for diff from Chance: 3.956
(.05 requires 1.96)
(.01 requires 2.58)
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .05, the critical ratio will exceed 1.96.
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .01, the critical ratio will exceed 2.58.
A chance probability of .05 or less is sufficient to indicate bias in the lineup.
A rule of thumb is that no more than 50 mock witnesses should be used.
Confidence intervals
5% confidence interval equals 0.385 to 0.815
1% confidence interval equals 0.317 to 0.883
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 5% confidence interval, we can be 95% certain that the
lineup is biased.
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 1% confidence interval, we can be 99% certain that the
lineup is biased.
My Results using the Lineup bias calculator:
Instructions.
1. Enter the total number of mock-witness choices here: 12
2. Enter the number of mock witnesses who chose the suspect here:1
3. Enter the total number of persons in the lineup here:6
Note:Fields in which data can be entered are shown in the boxes.Other fields are calculated fields.
Exact probability of these results 0.269176
Proportion identifying the suspect (p) .083
q = 1 - p .917
n 12
Standard Error of p .080
Chance expectation: .167
Critical Ratio for diff from Chance: -1.044
(.05 requires 1.96)
(.01 requires 2.58)
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .05, the critical ratio will exceed 1.96.
If the proportion of mock witnesses identifying the suspect reveals bias,
beyond a chance probability of .01, the critical ratio will exceed 2.58.
A chance probability of .05 or less is sufficient to indicate bias in the lineup.
A rule of thumb is that no more than 50 mock witnesses should be used.
Confidence intervals
5% confidence interval equals -0.073 to 0.240
1% confidence interval equals -0.123 to 0.289
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 5% confidence interval, we can be 95% certain that the
lineup is biased.
If the proportion of mock witnesses expected to identify the suspect by chance
is outside of the 1% confidence interval, we can be 99% certain that the
lineup is biased.
Although only one witness out of 12 chose the suspect, my lineup was still biased, because 6 of them still chose one particular filler. For the lineup to not be biased, there should not be a large number of witnesses choosing certain people.
Lindsey Fails
I thought this was pretty neat. I really didn’t know the bias would be so strong. I suppose the human mind picks up on relatively easy annotations therefore line-ups need to be very similar yet not to close to where a victim or eyewitness can’t tell who the perpetrator is at all. Searching through the database was interesting finding people that looked similar yet not exactly the same. I did however put a fairly good lineup together, gave a vague description, and I feel that my line up is fairly un-biased. Here are my results:
10 witnesses
2 chose the suspect
6 in the lineup
Probability= .290710
Proportion=.2
Standard error=.126
Chance expectation=.167
Crit ratio= .264
Distribution was fairly equal, at least one of each of my lineup guys were chosen at least once, so from that I think I came up with a pretty good lineup with a brief, general description of the perp. I really liked the access of the Florida database---yet I can see how easy it would be to give an unbiased description as well as putting biased pictures together.
I had a lot of fun with this assignment. It was a challenge to find 6 men that were similar in characteristics, but not too similar. I took a lot of time to work on this and try to create what I thought was a fair line up. My biggest problem was that my suspect was 20, and looked very young, and the others who fit the description, and were the same age, looked a lot older. I think this is why most people chose the suspect, but 6/14 isnt that many. My findings were
Exact probability of these results 0.014969
Proportion identifying the suspect (p) .429
q = 1 - p .571
n 14
Standard Error of p .132
Chance expectation: .167
Critical Ratio for diff from Chance: 1.980
(.05 requires 1.96)
(.01 requires 2.58)
I thought this mock witness line-up was very interesting and revealed how bias and unreliable eyewitness identification can be. Unless the description of the suspect is perfect, the suspect may easily be mistaken for, a filler in the line-up. I described the suspect as a White male with short brown hair and blue or green eyes, thin build, with a weight of about 165 pounds standing just under 6 feet tall. After a long search I found 5 fillers to mix in with my suspect that looked similar but not too similar to my suspect.
1. Enter the total number of mock-witness choices here: 15
2. Enter the number of mock witnesses who chose the suspect here: 6
3. Enter the total number of persons in the lineup here: 6
Note: Fields in which data can be entered are shown in boxes.
Other fields are calculated fields.
Exact probability of these results 0.020791
Proportion identifying the suspect (p) .400
q = 1 - p .600
n 15
Standard Error of p .126
Chance expectation: .167
Critical Ratio for diff from Chance: 1.845
(.05 requires 1.96)
(.01 requires 2.58)
I figured my results would come out similar to how they did. Of 15 mock witnesses only 6 chose the correct person out of the 6 members of the line-up, 40%, a fairly low probability of accurate results. The description is the key to positive identification, the more details you give out the easier it is to pick the correct suspect. This activity was pretty difficult but very informative; I have a new respect for the experts that have to do this on a daily basis.