This present study examines the effect of suspicion on personality ratings. One Hundred subjects were selected at random from the Victoria Police (Male=81 Female=19). No specific selection criteria was employed. Ten digitised photographs of five Doctors plus five of known convicted violent criminals were presented to the subjects.The identities of the photographic subjects shown, were not made known to the subjects throughout the questionnaire. Using four characteristics of personality, within semantic differential scales, Police subjects rated the photographic subjects, (this was the initial observation). The Police subjects thoughts were then re-focussed by them undertaking an assessment of their perception of the personality characteristics of doctors and criminals using the same scales (the suspicion phase), this was conducted without the aid of any photographs. The Police subjects then re-assessed the same initial ten photographic subjects with the suspicion that these photographic subjects now represented either doctors or criminals. Results would indicate that suspicion alone does affect personality ratings, based purely on facial appearance and no other cues, but that the experience or the age of the Police subjects did in fact contribute to any bias presented.
EFFECTS OF SUSPICION ON PERSONALITY RATINGS.
When a person meets or views another person for the first time, what often comes to the observers mind is that of an image of the typical member or prototype of the group that the observed may or may not belong to. The observed exhibits a personality, a characteristic or a recognisable pattern of traits that are known or recognised by the observer. Once a person has been associated with such a type or category of person that the observer considers that the person fits, they are automatically assumed to possess many of the same characteristics as the typical group member (Cantor & Mischel, 1979). Such person perception judgement could be seriously distorted by such a characterisation.
In any such face to face interaction, a person observes anothers verbal and non-verbal behaviours. they encode these behaviours into some form of representation, usually a combination of images and verbal descriptions. an example by Gergen, 1982, was, David suddenly moved his clenched fist forward and the fist connected with the mans chin. The actual description would probably be, David punched the man. Thus by example, when told that David punched the man, it can be inferred that Davids actions were intentional and that he was possibly an aggressive person (Smith & Miller, 1983). Furthermore, one can conclude that people will not like David because of his aggressive behaviour.
No thought or reason is required to make such a judgement, only that the person receiving such information had through past experience and social learning deduced that David through using violence must own those same traits, those of aggression.
A person is thus categorised because of an action seen or perceived by others, or recalled from long term memory. If a person takes up rock climbing, would they then be called adventurous or reckless? It is this use of categorisation that of adventurous or reckless that is retrieved from long term memory which completes this categorisation of the person into either that of being reckless or adventurous (Martin, 1986). In folklore it is said that first impressions count. The information that is first obtained when a person first observes another person does appear to influence the overall impression of them much more than information obtained about the observed at a later time, this is known as the Primacy effect. (Asche, 1946). Primacy effects can be reduced however under three conditions; Passage of time, if a substantial period of time passes between first and subsequent impressions, the more recent impression will be of greater importance. This could mean that if a person is unfavourably viewed at the first encounter then with a second chance may be in fact seen in a more favourable light. (eg. At a personal interview, the first attempt saw the subject ill at ease, but with time and a second attempt the subject appears more relaxed, in control because of prior knowledge of the interview strategies).
Prolonged exposure to a person tends to reduce the importance of the first impression to others. Information gathered over a long period of time will erase any first impressions. Knowledge of primacy effects are when people are warned to avoid being influenced by first impressions, the primacy effect can be reduced (Hovland, 1957).
This is a situation where such people as personnel managers and others to whom accurate person perception is important are educated to the dangers of primacy effects and are often able to eliminate them from their perceptions of the subject. As more is discovered about a person across a number of social events, (eg. workplace, home etc) it is often discovered that during these processes a number of incompatible traits exist about the observed. Information is often discounted that is inconsistent with the initial impressions gained (Devine & Ostram, 1985). Forgas, 1985 stated that impressions about people can also be base upon information about a group of people. The ability to correctly recognise the characteristics of the group that a person belongs to, can greatly assist in the process of impression formation. These characteristics must then be seen as common traits of the whole group to be of assistance in the perception of how good or bad a persons personality actually is. Groups may have common goals or beliefs, these may be goals or beliefs that each individual member must hold to gain membership. The individual persons character within the group may differ, but their beliefs may be the same.
This then can be seen as a first impressions count theory. Making assumptions that because the group norm is, then the person is. Many characteristics or personality traits known by a person can attributed to people that they meet, observe or speak to. If the decision is to like or dislike a person, this combination of trait and other information is used to form a total impression.
Asche, 1946 simulated a situation by presenting subjects (students) with one of two lists of trait adjectives that described a particular person. Students receiving the first list were told that the person was, intelligent, skilful, industrious, warm, determined, practical and cautious. subjects receiving the second list were told that the person was, intelligent, skilful, industrious, cold, determined, practical and cautious. Note that both lists were identical except for the word warm in the first list and cold in the second. Despite the high similarity between the two lists, 90% of the students given the list containing the word warm thought that the person would also be generous, happy and sociable. Less that 40% thought that the cold person would display these qualities. Asches study illustrates how various traits are related to each other, and that when given pieces of information about a person, the information is integrated into coherent and unified impressions. Even though six of the seven pieces of information about the two people were identical, the overall impressions formed were quite different.
Asche also concluded that a persons powers of perception are a constructive rather than a passive process where the perceivers prior knowledge of what people are like plays an important role in determining their perceptions of others. An example of this would be the student who listens attentively to a tutor who is considered to be warm and friendly, the student will be laughing at jokes told by the tutor, nodding in agreement to points made, and generally responding in a positive manner towards the tutor. In fact the student is returning the perceived warmness, and the tutor could perceive this from the student and an excellent relationship could be formed. This all because a fellow student may have remarked how friendly and warm the tutor was when asked by the first student. This is called self fulfilling prophesies. The student was told of the warmness, expected it, and after meeting the tutor returned the warmth. This describes the phenomenon in which the initially false expectancies (prophesies) become confirmed (fulfilled) because of their effect on the interaction sequence between the two people.
Snyder, Tanke & Berscheid (1977) showed male subjects a picture of a woman with whom they believed they were going to have a telephone conversation with. Half of the subjects were shown a very attractive woman and the other half, an unattractive one. The women they actually talked to were both of average attractiveness. the research was to see how the beliefs of the males would affect they way the women responded during the telephone conversations.
Analysis of the womens comments during the conversations revealed that the women who were presumed attractive did in fact speak in a more warm likeable way than the women who were presumed unattractive. The mens mistaken beliefs had become a self-fulfilling prophesy, leading them to influence the women into fulfilling the stereotype that attractive people are desirable people.
People do tend to evaluate the qualities of others purely on physical appearance, (attractiveness), Dion, 1972 stated, what is beautiful is good. Positive attributes are then placed on a par with being attractive. Ratings of qualities from initial impressions would also be influenced by giving a warm/positive or cold/negative description of a subject, (Kelley, 1950). therefore if a person is observed by another as attractive, instead of cold, how much more of that person would one know of them? The adage, beauty is only skin deep may be the norm, but people may also try to interpret whether the subject is friendly or hostile, deceitful or truthful, intelligent or unintelligent? These interpretations or guesses are what most people would say; the person will more likely be intelligent than unintelligent, friendly more than hostile, truthful more than deceitful (Bernscheid & Walste, 1978). This inference that a persons physical appearance gives an insight into the other persons character. It is seen when children are taught at an early age that beauty is good, ugly is evil. People also theorise that personality characteristics are associated with specific physical attributes such as, overweight, thin, bearded are often found wanting.
Knowing that a man has a beard may suggest to some that he intelligent and strong, but not much more. the man could be of a religious order, or he could be a member of an outlaw motorcycle gang (Roll & Verinis, 1971). Social categorisation is where inference is drawn of a persons personality by such things as gender, race, occupation and nationality to document a few. The stereotypes that children inherit are shown by a study by Dion, 1972 that people viewed the minor misbehaviour of an attractive child as less naughty than the same actions performed by an unattractive child. Stewart, 1980 found that attractive defendants in serious criminal trials received shorter sentences than unattractive defendants. In regard to people being stereotyped as criminals; eye-witness testimony in police identification parades (line- ups), police photo books (mug-shots) have been seen to require research in an effort to evaluate eyewitness accounts of crime. Eyewitness identifications have led to the false imprisonment of innocent persons. In 1975 two men were convicted of kidnap, despite their pleas of innocence. Both men received sentences of from 28 to 40 years in prison. The convictions were largely gained on the eyewitness testimony of the kidnapped person. This person had only a short glimpse of the offenders before they donned ski masks. Investigators took two years to find the actual two male offenders, both original men were pardoned and released. On being asked why he had identified the original two males, the victim stated, they looked like them, and I thought they looked bad enough, and the cops line-up wouldnt have innocent men in it! (Loftus, 1979).
This case concedes that eyewitness testimony is dangerous when presented alone. It also gives rise to the myth that police line-ups, photo books only contain criminals and that police expect a positive identification from either. In Victoria, persons required to attend a line-up now can view the parade via video screen. They are informed that the parade is not all criminals, also the fact is that some parades do not even contain the alleged offender. This places less pressure on the witness to come up with the goods.
Victoria police have also attended to the mug shots by now using a collage of photographs that contain up to twelve shots of persons, off the street. The offender is allowed to place his/her photo anywhere into the collage. He/She can even object to other subjects with the line-up or collage that do not look like the alleged offender. This makes the procedure more fair to the offender and the witness. It can be imagined if a witness tells a jury that the man in the dock is the offender, this base purely on a fleeting glance, but also on a preconceived notion that he must be bad to be under police notice. The jury may hold the same view as that of the witness, they then perceived the person in the dock and make first impression judgement. this does not appear to be fair to the person in the dock. That person could be found guilty of a crime that they did not commit! Then one must consider, a person informs another that a third party is a criminal (bad), there are grounds to believe that the perception of that person will be influenced.
This requirement to view or judge a person based solely on a first impression without any other cue is one that such people as law enforcement personnel are required to make every working day. The officer is required to make an instant decision to stop and detain another person for questioning, this base purely on a perception that the subject is, or about to commit a crime. Do police officers have pre-conceived ideas like the eyewitness? Evidence would suggest that the police officer is no different than any other in regards to social biases. An officer is taught to be suspicious, taught to question thoroughly before making a snap decision. In reality how good is the training to identify the vital cues that point to a wrong-doer. Does an officers training, age or gender change their perception of good or bad?
This study will address;
Doctor/criminal types were selected as socially representing the two extremes of what is publicly seen as good and bad. The assessment by the subjects of rating for certain personality characteristics is thought to focus their thoughts towards what is considered to be good and bad personality characteristics.
METHOD
Subjects
One Hundred persons (Male N=81; Female N=19) took part in the research, all serving members of the Victoria Police Force. The age of the subjects ranged from 25 years to 50 years. The subjects years of service in the Police Force ranged from three years to thirty one years. The subjects were utilised without any preselection criteria. The personnel were representatives of country, suburban, and specialist squads. The majority being obtained from within the uniform section of the police department.
Apparatus
Ten digitised photographs of Health professionals (Doctors/Psychologists) were constructed from photographs taken from readily obtainable text books.
The ten digitised photographs of the criminals were taken from the autobiography of a well known criminal, and from actual public records. All context cues ( background, surrounds and dress) were removed during the computer digitising process.
A set of semantic differential scales consisting of four ratings, all rated from one to ten, was devised. These scales were;
a. Not Intelligent to Intelligent. b. Dangerous to Not Dangerous. c. Cruel to Caring. d. Hostile to Friendly.
Procedure
A pre-trial of the personality characteristics was conducted prior to the actual experiment. Sixty members of the police department were requested to describe what characteristics or personal traits that they would expect to be exhibited by the doctor (health professionals) and the criminal groups. A construction of a list of personality ratings that consisted of forty descriptors was obtained from these members of the Police force. This was reduced until four ratings remained as a result of a further series of small trials.
A full pilot study of ten police subjects was then conducted using the actual research questionnaire to trial the scales to be used. As a result of this mini trial, the original research which was to include twenty two photographs, was reduced to ten photographs. This due to subjects interest beginning to fade due to the length of time required to complete such a questionnaire, and the time restraints attached to the research.
The actual questionnaire consisted of five sections; (Appendix A.).
A page consisting of questions relating to approximate age, gender and length of service.
Annex A: Consisted of the ten digitised photographs with an opposite page displaying the four semantic scales to be completed. Subjects were invited to view each photographic subject, then to rate that subject using the semantic scales provided opposite. At no time during this procedure were the subjects made aware of the actual status of the photographic subjects being a Doctor or a Criminal.
Annex B: Using a new set of the four semantic scales, subjects were required to rate what they considered was in their view a definition of a Doctor or Medical/Health professional.
Annex C: Using a new set of the four semantic scales, subjects were required to rate what they considered was a violent criminal.
NOTE: These were the two sections (Annex B & C) that were to implant the suspicion into the minds of the subjects that the photographic subjects were either Doctor or Criminal.
Annex D: A further set of the same ten digitised photographs was then re-presented to the subjects. These photographic subjects were to be re-rated using the same semantic differential scales as previous. Research subjects were still not informed of which photograph depicted a Doctor or a Criminal. At the conclusion of the research questionnaire, the subject was given the opportunity to discover which photograph was in fact a Doctor or a criminal.
The subjects had then rated the photographs on face value or first impression. They had then had a suspicion sown in their minds that the photographic subjects were in fact either Doctors or Criminals by asking them to give a rating on Doctors and Criminals. Finally the subjects were required to re-rate the photographic subjects to establish if in fact they had changed their impressions once suspicion had been aroused by Annex B & C.
A method of abbreviation was used to describe the four semantic differential scales, Intell, Danger, Caring and Friendly throughout this research. These abbreviations are found detailed as Table 1.
Table 1. Description of abbreviated forms used to represent semantic scales.
Descriptor Actual Semantic Differential Scale. (Range 1 to 10).
(example: Not Intelligent 1 2 3 4 5 6 7 8 9 10 Intelligent.)
Pre-Test: Intell. Pre-Suspicion: Not Intelligent to Intelligent.
Pre-Test: Danger. Pre-Suspicion: Dangerous to Not Dangerous.
Pre-Test: Caring. Post-Suspicion: Cruel to Caring.
Pre-Test: Friendly. Pre-Suspicion: Hostile to Friendly.
Post-Test: Intell. Post-Suspicion: Not Intelligent to Intelligent.
Post-Test: Danger. Post-Suspicion: Dangerous to Not Dangerous.
Post-Test: Caring. Post-Suspicion: Cruel to Caring.
Post-Test: Friendly. Post-Suspicion: Hostile to Friendly.
Subjects, age, gender and years of service, were given a descriptor to divide the age of subjects into eight age groups, to divide years of service into three groups, and gender was divided into two groups (Table 2.)
Table 2. Description of the abbreviated method used in the division of subject groups.
Approximate Age of Subjects Years of Service in Force Gender of subject Descriptor Age Group Descriptor Years Served Descriptor Gender 1. 25 < years. Young 4-8 years. 1. Male. 2. 25-30 years. Middle 10-15 years. 2. Female. 3. 30-35 years. Mature 20-31 years. 4. 35-40 years. 5. 40-45 years. 6. 45-50 years 7. 50-55 years. 8. 55 > years.
RESULTS
On completion of the questionnaires, all data was entered into a data file using a personal computer which produced standard ASCII files. This was the preliminary merging of the data for analysis. The ASCII files were then transformed into the necessary variables to allow for analysis using an SPSSWIN statistical package. The variables used in the data analysis are found in Table 1. These variables were further categorised into subsets of variables, (Table 2). The Pre-Test, Post-Test scores of all four semantic differential scales were summed for each of the series of ten photographs. Tables 3a and 3b show the summed means and standard deviations for Pre-Test and Post-Test respectively for the four semantic differential scales. The raw statistical data is contained in Appendix B.
Table 3a. Pre-Test Means and Standard Deviations for the Four Semantic Differential Scales.
Photo N Intelligence Danger Caring Friendly Number value M SD M SD M SD M SD 1. 100 6.85 1.23 7.69 1.04 7.77 1.04 7.92 1.04 2. 100 6.24 1.25 4.98 1.27 4.67 1.18 4.74 1.15 3. 100 6.00 1.09 6.01 1.42 6.07 1.27 6.13 1.35 4. 100 6.19 1.22 4.90 1.30 4.64 1.11 4.60 1.19 5. 100 7.01 1.22 5.77 1.28 5.59 1.25 5.60 1.23 6. 100 5.26 1.12 4.57 .93 4.37 .86 4.39 1.05 7. 100 7.86 1.18 6.71 1.56 6.62 1.57 6.61 1.48 8. 100 7.67 1.46 6.80 1.49 6.69 1.51 6.75 1.45 9. 100 7.90 1.20 6.16 1.69 5.94 1.63 6.00 1.60 10. 100 5.43 1.77 3.56 1.36 3.26 1.21 3.33 1.25
Table 3b. Post-Test Means and Standard Deviations for the Four Semantic Differential Scales.
Photo N Intelligence Danger Caring Friendly
Number value M SD M SD M SD M SD
1. 100 7.48 1.15 8.14 1.07 8.12 1.04 8.17 1.02
2. 100 6.56 1.32 5.09 1.44 4.85 1.28 4.85 1.31
3. 100 6.18 1.18 6.37 1.34 6.27 1.31 6.46 1.23
4. 100 6.24 1.14 4.92 1.18 4.72 1.19 4.60 1.09
5. 100 7.26 1.15 5.99 1.51 5.94 1.54 5.85 1.51
6. 100 5.18 1.21 4.51 1.01 4.30 .97 4.20 .94
7. 100 7.89 1.35 7.04 1.49 7.05 1.45 6.95 1.49
8. 100 7.94 1.25 6.80 1.61 6.88 1.66 6.92 1.45
9. 100 7.88 1.34 6.41 1.58 6.27 1.67 6.33 1.56
10. 100 5.85 1.78 3.64 1.46 3.31 1.27 3.23 1.23
Table four shows the summed mean scores and standard deviations of the four semantic differential scales for Pre-Test of All Subjects compared with the summed mean scores of the four semantic differential scales for Pre-Test for subjects within the 25 to 30 Year Age group.
Pre-Test Semantic Scales All Subjects Group 25 - 30 Year Age Group
Mean Std Dev N Mean Std Dev N
Pre-Test Intell- 6.64 .97 100 6.86 1.07 17
Pre-Test Danger- 5.72 1.23 100 6.13 1.15 17
Pre-Test Caring- 5.56 1.33 100 5.81 1.43 17
Pre-Test Friendly- 5.61 1.35 100 5.78 1.39 17
The results of Pre-Test, Not Intelligent-Intelligent scale for All Subjects shows a mean score of 6.64 (SD= .97) and for 25-30 Year Age group was 6.86 (SD= 1.07). With alpha set at 0.05, a paired t-test showed a significant difference between All Subjects and 25-30 Year Age group subjects (t = -4.70, df = 9, p = .001). Thus it could be concluded that the Pre-Test 25-30 Year Age group was more positive in their perception of the Intelligence of the photographic subjects than the perceptions of the All Subjects group.
The results of Pre-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.72 (SD = 1.22) and for 25-30 Year Age group was 6.13 (SD = 1.15). With alpha set at 0.05, a paired t-test showed a significant difference between All Subjects and 25-30 Year Age group subjects (t = -2.43, df = 9, p = .038). Therefore this could be concluded that the 25-30 Year Age group subjects perceived the photographic subjects as less dangerous than the perception by the All Subjects group.
The results of Pre-Test Cruel-Caring for all subjects shows a mean score of 5.56 (SD = 1.33) and for 25-30 Year Age group was 5.81 (SD = 1.42). With alpha set at 0.05, a paired t- test showed a significant difference between All Subjects and the 25-30 Year Age group. (t = -3.25, df = 9, p = .010). This could be concluded that the 25-30 Year Age group perceive the photographic subjects as more caring than the perception of the photographic subjects by the All Subjects group.
The results of Pre-Test Hostile-Friendly for All Subjects group shows a mean score of 5.60 (SD = 1.35) and for the 25-30 Year Age group was 5.77 (SD = 1.39). With alpha set at 0.05, a paired t-test found no significant difference in the Hostile-Friendly semantic scale. (t = -2.02, df = 9, p = .074).
Table 5 shows the summed mean scores and standard deviations of the four semantic differential scales for Pre-Test of All Subjects compared with the summed mean scores and standard deviations of the four semantic differential scales for Pre-Test for subjects in the 25-30 Year Age group.
Post-Test Semantic Scales All Subjects Group 25-30 Year Age Group
Mean Std Dev N Mean Std Std Dev N
Post-Test Intell- 6.85 .98 100 6.98 .99 17
Post-Test Danger- 5.89 1.34 100 5.98 1.38 17
Post-Test Caring- 5.77 1.46 100 5.88 1.45 17
Post-Test Friendly- 5.76 1.51 100 5.85 1.46 17
The result of Post-Test Not Intelligent - Intelligent for All Subjects shows a mean score of 6.85 (SD = .98) and for 25-30 Year Age group was 6.98 (SD = .99). With alpha set at 0.05, a paired t-test found no significant difference in the Not Intelligent - Intelligent semantic scale (t = -1.23, df = 9, p = .250).
The results of Post-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.89 (SD = 1.34) and for 25-30 Year Age group was 5.98 (SD = 1.38). With alpha set at 0.05, a paired t-test found no significant difference in the Dangerous - Not Dangerous semantic scale. (t = -.80, df = 9, p = .446). The results of Post-Test Cruel - Caring for All Subjects shows a mean score of 5.77 (SD = 1.45) and for 25-30 Year Age group was 5.87 (SD = 1.45). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scale. (t = -.93, df = 9, p = .378). Table six shows the summed mean scores and standard deviations Pre-Test scores for All Subjects group compare with the summed mean scores and standard deviations Pre-Test scores for 30-35 Year Age group.
Pre-Test Semantic All Subjects Group 30-35 Year Age Group Scales Mean Std Dev N Mean Std Dev N Pre-Test Intell- 6.64 .97 100 6.53 1.13 35 Pre-Test Danger- 5.72 1.23 100 5.57 1.31 35 Pre-Test Caring- 5.56 1.33 100 5.34 1.36 35 Pre-Test Friendly- 5.61 1.35 100 5.46 1.44 35
The results of Pre-Test Not Intelligent - Intelligent for All Subjects shows a mean score of 6.64 (SD = .97) and for 30-35 Year Age group was 6.53 (SD = 1.13). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scale (t = 6.66, df = 9, p = .132). The results of Pre-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.71 (SD = 1.22) and for 30-35 Year Age group was 5.57 (SD = 1.30). With alpha set at 0.05, a paired t-test showed a significant difference between All subjects and 30-35 Year Age group. (t = 2.92, df = 9, p = .017). this could be concluded that the 30-35 Year Age group perceived the photographic subjects as more dangerous than did the All Subjects group. The results of Pre-Test Cruel - Caring for All Subjects shows a mean score of 5.56 (SD = 1.33) and for 30-35 Year Age group was 5.33 (SD = 1.35). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scale. (t = 2.01, df = 9, p = .075). The results of the Pre-Test Hostile - Friendly for All Subjects shows a mean score of 5.60 (SD = 1.35) and for 30-35 Year Age group was 5.45 (SD = 1.44). With alpha set at 0.05, a paired t-test showed a significant difference between All subjects and 30-35 Year Age group. (t = 3.53, df = 9, p = .006). Therefore it could be concluded that the 30-35 Year Age group perceived the photographic subjects as more hostile than did the All Subjects group.
Table seven shows the summed mean scores and standard deviations of the four semantic differential scales for Post-Test of All Subjects compared with the summed mean scores and standard deviations of the four semantic differential scales for Post-Test 30-35 Year Age group.
Post-Test Semantic Scales All Subjects Group 30-35 Year Age Group
Mean Std Dev N Mean Std Dev N
Post-Test Intell- 6.85 .98 100 6.82 1.13 35
Post-Test Danger- 5.89 1.34 100 5.89 1.45 35
Post-Test Caring- 5.77 1.46 100 5.76 1.61 35
Post-Test Friendly- 5.76 1.51 100 5.73 1.65 35
The results of Post-Test Not Intelligent - Intelligent for All Subjects shows a mean score of 6.85 (SD = .98) and for 30-35 Year Age group was 6.82 (SD = 1.13). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scale. (t = .29, df = 9, p = .775). The results of Post-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.89 (SD = 1.34) and for 30-35 Year Age group was 5.89 (SD = 1.44). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scale. (t = .00, df = 9, p = 1.00).
The results of Post-Test Cruel - Caring for All Subjects shows a mean score of 5.77 (SD = 1.45) and for 30-35 Year Age group was 5.76 (SD = 1.60). With alpha set at 0.05, a paired t- test showed no significant difference with the Cruel - Caring semantic scale.(t = .11, df = 9, p = ..911). The results of Post - Test Hostile - Friendly for All Subjects shows a mean score of 5.75 (SD = 1.50) and for 30-35 Year Age group was 5.73 (SD = 1.64). With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scale. (t = .37, df = 9, p = .717).
Table eight shows the summed mean scores and standard deviations of the four semantic differential scales for Pre-Test of All Subjects compared with the summed mean scores and standard deviations of the four semantic differential scales for Pre-Test for subjects in the 35 - 40 Year Age group.
Pre-Test Semantic Scales All Subjects Group 35-40 Age Group
Mean Std Dev N Mean Std Dev N
Pre-Test Intell- 6.64 .97 100 6.55 .84 18
Pre-Test Danger- 5.72 1.23 100 5.77 1.21 18
Pre-Test Caring- 5.56 1.33 100 5.62 1.36 18
Pre-Test Friendly- 5.61 1.35 100 5.67 1.33 18
The results of Pre-Test Not Intelligent - Intelligent for All Subjects shows a mean score of 6.64 (SD = .97) and for 35-40 Year Age group was 6.55 (SD = .84). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scale. (t = .82, df = 9, p = .435). The results of Pre-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.71 (SD = 1.22) and for 35-40 Year Age group was 5.77 (SD = 1.21). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scale. (t = -.68, df = 9, p = .513). The results of Pre-Test Cruel - Caring for All Subjects shows a mean score of 5.62 (SD = 1.34) and for 35-40 Year Age group was 5.62 (SD = 1.36). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scales. (t = -.65, df = 9, p = .531). The results of Pre-Test Hostile - Friendly for All Subjects shows a mean score of 5.61 (SD = 1.36) and for 35- 40 Year Age group was 5.67 (SD = 1.33). With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scales. (t = -.73, df = 9, p = .482).
Table nine shows the summed mean scores and standard deviations Post -Test scores for All Subjects group compared with 35-40 Year Age group.
Post-Test Semantic Scales All Subjects Group 35-40 Year Age Group
Mean Std Dev N Mean Std Dev N
Post-Test Intell- 6.85 .98 100 6.77 .933 18
Post-Test Danger- 5.89 1.34 100 5.85 1.30 18
Post-Test Caring- 5.77 1.46 100 5.71 1.36 18
Post-Test Friendly- 5.76 1.51 100 5.69 1.50 18
The results of Post-Test Not Intelligent - Intelligent for All Subjects shows a mean score of 6.85 (SD = .98) and for 35-40 Year Age group was 6.77 (SD = .933). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scales. (t = .91, df = 9, p = .385). The results of Post-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.89 (SD = 1.35) and for 35-40 Year Age group was 5.85 (SD = 1.30). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scales. (t = .41, df = 9, p = .695). The results of Post-Test Cruel - Caring for All Subjects shows a mean score of 5.77 (SD = 1.56) and for 35-40 Year Age group was 5.71 (SD = 1.36). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scales. (t = .56, df = 9, p = .592).
The results of Post-Test Hostile - Friendly for All Subjects shows a mean score of 5.76 (SD = 1.50 and for 35-40 Year Age group was 5.69 (SD = 1.50). With alpha set at 0.05, a paired t-test showed no significant difference with Hostile - Friendly semantic scales. ( t = .59, df = 9, p = .571).
Table ten shows the summed mean scores and standard deviations of the four semantic differential scales for Pre-Test of All Subjects compared with the summed mean scores and standard deviations of the four semantic differential scales in the 4-8 Years of Service group.
Pre-Test Semantic Scales All Subjects Group 4 - 8 Years of Service Group
Mean Std Dev N Mean Std Dev N
Pre-Test Intell- 6.64 .97 100 6.82 1.10 24
Pre-Test Danger- 5.72 1.23 100 5.81 1.30 24
Pre-Test Caring- 5.56 1.33 100 5.64 1.47 24
Pre-Test Friendly- 5.61 1.35 100 5.63 1.45 24
The results of Pre-Test Not Intelligent - Intelligent for All Subjects shows a mean score of 6.64 (SD = .97) and for 4-8 Years of Service group 6.82 (SD = 1.10).
With alpha set at 0.05, a paired t-test showed a significant difference between All subjects and 4-8 Years of Service subjects. (t = -3.68, df = 9, p = .005). This could be concluded that the 4-8 Years of Service subjects perceived the photographic subjects to be more intelligent than did the All Subjects group. The results of Pre-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.72 (SD= 1.22) and for 4-8 Years of Service subjects was 5.81 (SD = 1.30). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scales. (t = -1.31, df = 9, p = .222). The results of Pre-Test Cruel - Caring for All Subjects shows a mean score of 5.56 (SD = 1.34) and for 4-8 Years of Service subjects was 5.64 (SD = 1.47). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scales. ( t= -1.09, df =m 9, p = .303). The results of Pre-Test Hostile - Friendly for All subjects shows a mean score of 5.61 (SD = 1.36) and for 4-8 Years of Service subjects was 5.63 (SD = 1.45). With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scales.(t = - .30, df = 9, p= .771).
Table eleven shows the summed mean scores of the four semantic differential scales for Post-Test of all subjects compared with the summed mean scores of the four semantic differential scales for Post-Test for subjects within the 4-8 years of service group.
Post-Test Semanitc Scales All Subjects Group 4-8 Years of Service Group Mean Std Dev N Mean Std Dev N Post-Test Intell- 6.85 .98 100 6.89 1.12 24 Post-Test Danger- 5.89 1.34 100 5.94 1.47 24 Post-Test Caring- 5.77 1.46 100 5.79 1.65 24 Post-Test Friendly- 5.76 1.51 100 5.76 1.63 24
The results of Post-Test Not Intelligent - Intelligent for All Subjects showed a mean score of 6.84 (SD = .98) and for 4-8 Years of Service subjects was 6.89 (SD = 1.12). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scales. (t = -.70, df = 9, p = .502). The results of Post-Test Dangerous - Not Dangerous for All Subjects showed a mean score of 5.89 (SD = 1.35) and for 4-8 Years of Service was 5.94 (SD = 1.47). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scales. (t = -.48, df = 9, p = .644). The results of Post-Test Cruel - Caring for All Subjects showed a mean score of 5.77 (SD = 1.45) and for 4-8 Years of Service subjects was 5.79 (SD = 1.65). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scales. (t = -.19, df = 9, p = .855). The results of Post-Test Hostile - Friendly for All Subjects showed a mean score of 5.75 (SD = 1.51) and for 4-8 Years of Service subjects was 5.76 (SD = 1.63).
With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scales. (t = -.08, df = 9, p = .941).
Table twelve shows the summed mean scores of the four semantic differential scales for Pre-Test of All Subjects compared with the summed mean scores of the four semantic scales for Pre-Test for subjects with 10-15 years of service.
Pre-Test Semantic Scales All Subjects Group 10-15 Years of Service Group
Mean Std Dev N Mean Std Dev N
Pre-Test Intell- 6.64 .97 100 6.50 .99 43
Pre-Test Danger- 5.72 1.23 100 5.64 1.19 43
Pre-Test Caring- 5.56 1.33 100 5.48 1.31 43
Pre-Test Friendly- 5.61 1.35 100 5.53 1.33 43
The results of Pre-Test Not Intelligent - Intelligent for All Subjects showed a mean score of 6.64 (SD = .97) and for 10-15 Years of Service subjects was 6.50 (SD = .99). With alpha set at 0.05, a paired t-test showed a significant difference between All Subjects and 10-15 Years of Service subjects. (t = 3.72, df = 9, p = .005).
This could be concluded that the 10-15 Years of Service subjects perceived the photographic subjects as less intelligent than the perception of the All Subjects of the same photographic subjects. The results of Pre-Test Dangerous - Not Dangerous for All Subjects showed a mean score of 5.92 (SD = 1.22) and for 10-15 Years of Service subjects was 5.64 (SD = 1.19). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scales. (t = 2.11, df = 9, p = .064). The results of Pre- Test Cruel - Caring for All Subjects showed a mean score of 5.56 (SD = 1.34) and for 10-15 Years of Service subjects was 5.48 (SD = 1.31). With alpha set at 0.05, a paired t-test showed a significant difference between All Subjects and the 10-15 Years of Service subjects. (t = 2.27, df = 9, p = .050). This could be concluded that the 10-15 Years of Service subjects perceived the photographic subjects as less caring than did the All Subjects group. The results of Pre- Test Hostile - Friendly for All Subjects showed a mean score of 5.61 (SD = 1.36) and for 10-15 Years of Service subjects was 5.53 (SD = 1.33). With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scales. (t = 2.02, df = 9, p = .074).
Table thirteen shows the summed mean scores of the four semantic differential scales for Post-Test of All Subjects compared with the summed mean scores of the four semantic differential scales for Post-Test for subjects with 10-15 Years of Service.
Post-Test semantic Scales All Subjects Group 10-15 Years of Service Group
Mean Std Dev N Mean Std Dev N
Post-Test Intell- 6.85 .98 100 6.82 .99 43
Post-Test Danger- 5.89 1.34 100 5.82 1.37 43
Post-Test Caring- 5.77 1.46 100 5.71 1.46 43
Post-Test Friendly- 5.76 1.51 100 5.69 1.54 43
The results of Post-Test Not Intelligent - Intelligent for All Subjects showed a mean score of 6.85 (SD = .98) and for 10-15 Years of Service subjects was 6.82 (SD = .99). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scales. (t = .38, df = 9, p = .253). The results of Post-Test Dangerous - Not Dangerous for All Subjects showed a mean score of 5.89 (SD = 1.35) and for 10-15 Years of Service subjects 5.82 (SD = 1.37). with alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scales. (t = 1.22, df = 9, p = .253). The results of Post - Test Cruel - Caring for All Subjects showed a mean score of 5.77 (SD = 1.56) and for 10-15 Years of Service subjects was 5.71 (SD = 1.46). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scales. (t = 1.26, df = 9, p = .240).
The results of Post-Test Hostile - Friendly for All Subjects showed a mean score of 5.76 (SD =1.51) and for 10-15 Years of Service subjects was 5.69 (SD =1.54). With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scales. (t = 1.95, df = 9, p = .083).
Table fourteen shows the summed mean scores and standard deviations Pre-Test scores for All Subjects group compared with the summed mean scores and standard deviations for 20- 31 Years of Service subjects.
Pre-Test Semantic Scales All Subjects Group 20-31 Years of Service Group
Mean Std Dev N Mean Std Dev N
Pre-Test Intell- 6.64 .97 100 6.52 .83 43
Pre-Test Danger- 5.72 1.23 100 5.60 1.20 43
Pre-Test Caring- 5.56 1.33 100 5.49 1.20 43
Pre-Test Friendly- 5.61 1.35 100 5.63 1.27 43
The results of Pre-Test Not Intelligent - Intelligent for All Subjects showed a mean score
of 6.64 (SD = .97) and for 20-31 Years of Service subjects was 6.52 (SD = .83). With alpha set
at 0.05, a paired t-test showed no significant difference in the Not Intelligent - Intelligent
semantic scales. (t = 1.07, df = 9, p = .312).
The results of Pre-Test Dangerous - Not Dangerous for All Subjects showed a mean score of 5.72 (SD = 1.23) and for 20-31 Years of Service subjects was 5.60 (SD = 1.20). With alpha set at 0.05, a paired t-test showed no significant difference in the Dangerous - Not Dangerous semantic scales. (t = 1.07, df = 9, p = .312). The results of Pre-Test Cruel - Caring for All Subjects showed a mean score of 5.56 and for 20-31 Years of Service subjects was 5.49 (SD = 1.20). With alpha set at 0.05, a paired t-test showed no significant difference in the Cruel - Caring semantic scales. (t = .87, df = 9, p = .407). The results of Pre-Test Hostile - Friendly for All Subjects showed a mean score of 5.61 (SD = 1.36) and for 20-31 Years of Service subjects was 5.63 (SD = 1.27). With alpha set at 0.05, a paired t-test showed no significant difference in the Hostile - Friendly semantic scales. (t = -.24, df = 9, p = .816).
Table fifteen shows the summed mean scores and standard deviations of the four semantic scales for Post-Test of All Subjects compared with the summed mean scores and standard deviations of the four semantic differential scales for Post-Test 20-31 Years of Service group.
Post-Test Semantic Scales All Subjects Group 20-31 Years of Service Group
Mean Std Dev N Mean Std Dev N
Post-Test Intell- 6.85 .98 100 6.68 1.06 43
Post-Test Danger- 5.89 1.34 100 5.83 1.32 43
Post-Test Caring- 5.77 1.46 100 5.77 1.41 43
Post-Test Friendly- 5.76 1.51 100 5.72 1.39 43
The results of Post-Test Not Intelligent - Intelligent for All subjects shows a mean score of 6.85 (SD = .98) and for 20-31 Years of Service subjects was 6.68 (SD = 1.06). With alpha set at 0.05, a paired t-test showed no significant difference with the Not Intelligent - Intelligent semantic scales. (t = 1.61, df = 9, p = .141). The results of Post-Test Dangerous - Not Dangerous for All Subjects shows a mean score of 5.89 (SD = 1.35) and for 20-31 Years of Service subjects was 5.83 (SD = 1.32). With alpha set at 0.05, a paired t-test showed no significant difference with the Dangerous - Not Dangerous semantic scales. (t = .78, df = 9, p = .456). The results of Post-Test Cruel - Caring for All subjects shows a mean score of 5.77 (SD = 1.46) and for 20-31 Years of Service subjects was 5.77 (SD =1.41). With alpha set at 0.05, a paired t-test showed no significant difference with the Cruel - Caring semantic scales. (t = .00, df = 9, p = .997).
The results of Post-Test Hostile - Friendly for All subjects shows a mean score of 5.76 (SD = 1.51) and for 20-31 Years of Service subjects was 5.72 (SD = 1.39). With alpha set at 0.05, a paired t-test showed no significant difference with the Hostile - Friendly semantic scales. (t = .29, df= 9, p = .778).
DISCUSSION
In discussing the outcomes of this research and the perceived implications, a salient point needs to be stressed. That is, that the number of subjects and the statistics utilised by the researcher do not fully explore the subject. With the data provided however, a comparison of Pre and Post suspicion (Test) scores via a comparison of means, standard deviations and paired t-tests has in the case of All Subjects (N=100) supported the hypothesis A. The Pre suspicion scores have in all cases with the exception of photographic subject six, risen in the Post- suspicion phase. the subject in photograph six, a Female, can be seen to be an extremely dark photograph, it also depicts a young lady who gives the perception of looking at the observer in a weird way (words used by two research subjects). The spurious nature of the photograph may however have been the cause of the result. The hypothesis A in relation to bias in assessment has been proven due to the significant difference between Pre and Post test ratings. The reason for such bias is of interest as no other contextual cues were presented with the photographs. Overall the rating that has given the most consistent scores is that of Intelligence, the rating has shown to be higher on Post-suspicion in All Subjects statistics. The rating would tend to illustrate that intelligence is not merely seen as the domain of academics, but that criminals can be intelligent also.
The three remaining scales, Danger, Caring and Friendly exhibited in the All Subjects statistics all have risen to varying degrees. This would indicate that the subjects suspicion has been aroused by the two Annexs of Doctor (Annex B) and Criminal (Annex C) prior to the Post-suspicion rating. Subjects aroused suspicion in the Post-suspicion phase may have lead them to search for more succinct visual cues, eg. eyes, lips or facial hair. This was evident in post discussion with the research subjects who referred to, thin lips, evil eyes and beards. these very similar to that of Dion, (1972) where attractiveness was said to be positive. The results of the All Subjects comparison with the 25-30 years of age subjects appears to show that the 25-30 year old subjects have a more positive perception of the photographic subjects. It could also denote that the present training of the younger police by the Victoria Police Project Beacon has been of benefit, this where members are receiving training in thinking before acting. If the comparison between the All Subjects and 30-35 year age group in the pre-test phase is scanned, it can be seen that this group exhibits the perception that the photographic subjects are more hostile, more dangerous, and in the post-test phase there is not much change from the original perception.
One could conclude that at this age the officer has been exposed to the stresses of police duties and has over time received negative input into their perception of good and bad, along with its attendant problems. The officer has also formed pre-conceived ideas about what he considers is a criminal. The term he looks a bad bastard was used by a number of research subjects, this in reference to particular photographic subjects. Also this would seem to be the time of service that these officer have reached a peak in their career, and contend to know it all. The logical follow-up from this age group is that of the 35-40 year age group which exhibited a more positive perception once again similar to that of the younger 25-30 year age group. This would seem to indicate that the older more mature officer has over time, mellowed in the perception of good and bad, through exposure and learning through contact with the criminal elements in society that not everyone is a criminal. Or like Hovland, 1957, the first impression phase of meeting with the criminal element has changed due to becoming informed of the personal attributes of criminals. In other words, the officer becomes more informed about the criminal element through exposure over time. The 4-8 years of service subject compared with All Subjects further indicates that the officer with limited street experience shows less likelihood of considering every person that they meet as a criminal. They exhibit less negativity that the Subjects in the All Subjects group. This factor could be due to the young inexperienced officer not wanting to make such a decision.
Again the 10-15 years of service subjects tended to be the most negative of the three years of service groups. similar to the corresponding 25-30 year age group subjects. These two categories suggest that an officer in that age group and with that experience exhibited the most negative bias toward the photographic subjects. The 20-31 years of service subjects exhibit like their younger counterparts a more relaxed perception. The results would thus indicate that training tends to eliminate an amount of bias. It would also seem to possibly show in other ways any negative perception of the public by the police officer. The police officer is not just a negative personality whom one sees only when in trouble, but can be seen as having the same frailties as other member of society. the research points to the fact that possibly re-training or some form of remedial training needs to be done on a regular basis to eliminate the negativity exhibited by the Middle group subjects.
Lombroso, 1946 attempted to use anthropomorphic measurements in an effort to identify criminals. This theory would not appear to be correct. In this research if one compares the photographic subjects with the scores given to each photo by the research subjects it would appear that the correct perception of each photographic subject was commonly wrong. and when one considers that every effort was made to reduce the visual cues presented, then only the anthropomorphic details basically remained to be used.
It is of interest that during the pilot study conducted of this research, comments of actual doctors and other persons were solicited. Most doctors stated that they looked for collars, ties, glasses or beards. Middle age persons also tended to perceived criminals with looking evil. Middled age females remarked on such items as, kind eyes, laugh lines and the thin cruel lips. Young males of 30-35 years tended to make more snap decisions than other groups.
The situation then, of a jury passing judgement on a person is brought to mind in relation to first impression counts theories. The idea of not allowing the jury to see the person in the dock until after evidence has been given and then only after the person has been found guilty would be of immense interest. Would the jurys perception or verdict change because of the defendants looks, because of their criminal history or would it remain unchanged ? Is this why people attending court as defendants shave off beards, wear suits rather than their usual attire or bring along their frail parents for sympathy. The case of the victim of a crime or the witness to an event attending a police station to view the mug shots. Why are they called mug shots, is every person in the books a criminal ? If so does that mean that all persons present in a line-up must also all be criminals. How is a member of the public to know who is who if the police officer cannot tell. Interesting questions for more detailed research !
The HTML conversion of this chapter was supported by the
National Institute of Justice/
National Criminal Justice Reference Service
Washington, D.C.
This is an archive page that is no longer being updated. It may contain outdated information and links may no longer function as originally intended.