McWeeny - 111311ng on the inside back cover Church lane...

Info icon This preview shows pages 1–7. Sign up to view the full content.

View Full Document Right Arrow Icon
Image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
Image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 4
Image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 6
Image of page 7
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: .. 111311ng on the inside back cover, Church lane. East Finchley. LondOn .)0l (:5 Psychology. Lancashire <21"1"3'_i'r-lr().’ag_i' please apply to the 7? i from Swtifi 2%: Zeitlingcr B.V.. 3c (Enter. Philadelphia. 01’ separate articles for private use .tctl articles for non-profit classroom ;e. or tat? internal or personal use of for 1' iraries and ' users registered 20.13 J .é vice. r L .cd that the base itrtef bdlCi‘l’i, Pei/“s. 01970. USA. trim not extend to other kinds of a or i'or promotional purposes. for 31's name appears on the first page of . hcidcr. permission to reprint may be Vianziger. 132;). Church Lane. East May. ingust and - . Road East. Leicester ' 00. us, . A1900. Single Copy 301"; Z\-‘1caciaum Avenue. , Mimi). Hit. ‘riui Hi i"$ViLllOl‘Li§f-' Publications nit}: J . J.) '1 Lili‘. (iCS‘lfltl‘w'iiL‘C, an. e the UK by .a.__d__av_‘a~——~———~V_.¢W kw-“ .__—-fi.y-—-1 Briti‘h Journal of Psychology (1987), 78, 143—149 Printed in Great Britain 143 © 1987 The British Psychological Society . Putting names to faces Kathryn H. McWeeny, Andrew W. Young, Dennis C. Hay and Andrew W. Ellis SubjeCtS were asked to learn occupations and surnames belonging to lo unfamiliar faces. The e,{peximent was intended to remove potential factors that might underlie the easrer recall of people’s occupations found in everyday life. Thus, contextual cues were eliminated by presenting faces in isolation, cueing from the retrieval of related items of semantic information was eliminated, and the frequency of use of occupational and name information was equated. Results showed that names . remained much harder to recall than occupations. This was true even for ambiguous labels that could be used as names or as occupations. It is much harder to recall that a person’s surname is Baker than to recall that a person is a baker. A common experience is to know exactly who a seen person is but be unable to bring to mind her or his name. When in this state we usually know the person’s occupation, where she or he is usually seen, who her or his friends are, and so on; only the name remains elusive (Young et al., 1985). In contrast to these recall difiiculties, recognition memory for names seems to be highly effective (Clarke, 1934; Bahrick et al., 1975). Problems in recalling people’s names produce tip-of-the-tongue 'states which have been studied in order to further understanding of the processes involved in the retrieval of items from memory (Gruneberg et al., 1973; Yarmey, 1973; Williams & Santos-Williams, 1980; Williams & Hollan, 1981; Read & Bruce, 1982; Reason & Mycielska, 1982; Reason & Lucas", 1984). What has received less attention, however, is the question as to why people’s names should be so difficult to recall (though a notable exception is the work of Morton £101., 1985). The'answer to this question is by no means obvious. Only a limited range of semantic characteristics, such as age or sex, can be reasonably accurately inferred from a person’s facial appearance. Such characteristics as a person’s occupation or the place where she or he is usually encountered are not closely linked to particular face shapes or certain types of facial feature. People do, of course, have occupational stereotypes concerning what, say, a‘typical’ film star or a ‘typical’ pop star might look like (Klatzky et al., 1982a, b), but thesecdo not provide veridical information. Thus, in the absence of contextual cues, kllotiing a seen person’s occupation can be as much a recall task as knowing that person’s name? ' DfiSpite this point, Hay & Young (1982) and Young et al. (1985) noted that whereas we m often remember a seen person’s occupation without being able to remember her or his name; we never seem to remember a seen person’s name without also remembering their 0Outpation. Hay & Young (1982), H. D. Ellis (1986), Young et al. (1985) and Bruce & fYIOllfi" 1986) used these, and other, observations to suggest that names are only accessed 0 is, then, a prima facie case for regarding names as relatively difficult things to recall Compared to other semantic information. Our interest in this paper is with why the hard to recall. Although it has been argued that names are relatively diflicult to ’ use they can only be accessed indirectly from familiar faces via intervening H 0 codes (Hay & Young, 1982; Young et al., 1985; Bruce & Young, 1986; His, 1986), it is easy to think of more prosaic potential explanations of the ©Il’0n. PSY 78 144 Kathryn H. McWeeny, A. W. Young, D. C. Hay and A. W. Ellis The most obvious possibility is that everyday life simply provides us with a wide Varie of non-facial contextual cues to occupation and the like. Thus we might know that Ms Aty :2: is a doctor because she is carrying a stethoscope, we might know that Mr B. works for the sec railways because he is checking our tickets, and so on. The errors studied by Young et a], the (1985) all came from descriptions of real-life incidents, in which the role of such contextual ' cues would be difficult to assess. H‘ It is also likely that, although we often need to recall them, names do not have to be as retrieved with anything like the frequency that other types of semantic information are :16 required. In addition, the retrieval of some items of semantic information may then one th p( retrieval of others. Thus, if you remember that a person you meet in town is someone you fr- usually see in a field it will be easier to remember that she or he is a local farmer. 16 Last, but by no means least, comes the point that names are often items of low meaningfulness or imageability (Cohen, Higgins, Jones, etc.) and therefore possibly 1: unmemorable for this reason alone. 11 Thus simple explanations as to why people’s names are retrieved less easily than other semantic information can be based on contextual cues, frequency of use, cues from other g semantic information, and low meaningfulness or imageability. All of these explanations imply that if such factors could be equated between names and other items of semantic information, then names would be no more difficult to retrieve than anything else. In , contrast, the explanation favoured by Young et a1. (1985) and Bruce & Young (1986), that names are only retrieved via intervening semantic codes specifying a person’s identity, would predict that a difference between ease of recall of names and other semantic information will continue to exist even when differences in cueing, frequency, and meaningfulness or imageability have been removed. The experiment reported here tests these contrasting predictions. The factors underlying the simple explanations of difficulties in name retrieval were eliminated by showing subjects '7 faces of unfamiliar people, and teaching them their (invented) names and occupations. Thus we could eliminate contextual cues by presenting each face in isolation, we could control the frequency of use of semantic or name information, and we could eliminate the role of cueing from other items of semantic information by teaching subjects only an occupation and a name for each person. To examine the effect of meaningfulness or imageability of individual items we made use of what we will call ambiguous labels; these have the property that they can be used as surnames or as occupations (Carpenter, Farmer, etc.). Method Subjects Sixteen university students (eight male, eight female) acted as subjects. All were paid. Stimuli and procedure Black~and—white photographs of the faces of 16 middle—aged men who were unfamiliar to the subjects were used as stimuli. These photographs were surrounded by a closely fitting circular template to remove as much clothing and background as possible, and rephotographed to a standard size to make the slides used in the experiment. Subjects were tested individually, and given spoken instructions by the experimenter. The 16 faces were shown to each subject one at a time, and s/he was told the person‘s (invented) surname and occupation. Half the subjects of each sex were told this information in the order surname then occupation, and half in the order occupation then surname. Thus half the subjects were told thatt ‘This man is called Mr X and he is a Y’, and half were told that ‘This man is a Y and he is called Mr X’. Subjects were asked to try to remember each person’s occupation and surname. . Each subject was then shown the 16 faces again, and was asked to state the person’s occupation 4. W. Ellis provides us with a wide variety hus we might know that Ms A, know that Mr B. works for the errors studied by Young et a]. Ihlch the role of such contextual m, names do not have to be of semantic information are 7 ,ic information may then cue the u meet in town is someone yOu 3r he is a local farmer. are often items of low .) and therefore possibly etrieved less easily than other [uency of use, cues from other lity. All of these explanations and other items of semantic ieve than anything else. In 1nd Bruce & Young (1986), that :cifying a person’s identity, mes and other semantic cueing, frequency, and lictions. The factors underlying : eliminated by showing subjects :d) names and occupations. 1 face in isolation, we could on, and we could eliminate the teaching subjects only an fect of meaningfulness or ill call ambiguous labels; these occupations (Carpenter, Farmer, ects. All were paid. «7—- r,_‘__~_n ¢._ WA.,,—v_____~,__n who were unfamiliar to the subjects )5er fitting circular template to ographed to a standard size to by the experimenter. The 16 faces L erson’s (invented) surname and )n in the order surname then half the subjects were told that This man is a Y and he is called ipation and surname. to state the person’s occupation Recalling names 145 and surname in the order in which s/he had learnt them. Any errors were noted by the experimenter, and corrected. This procedure was repeated, with 16 faces being presented in an unpredictable sequence for each repetition, until each subject had given correct occupations and surnames to all of the faces on two successive occasions. Surnames to be paired with the faces were chosen by searching the Lancaster, Morecambe and Heysham March 1984 Telephone Directory for surnames which could be used as either surnames or as occupations, such as Baker, Cook or Porter. These names which can be used either as surnames or occupations are referred to here as ambiguous labels. Sixteen of these ambiguous labels were chosen. There were then divided into two sets of eight ambiguous labels which were matched as closely as possible on frequency of occurrence as a surname in the Lancaster Telephone Directory, on frequency of occurrence as a word in the Kucera & Francis (1967) norms, and for the number of letters in the labels. A further eight surnames which could not be used as occupations (such as Hyde, Knowles, Rothwell) were then chosen to match the ambiguous labels on frequency of occurrence in the Lancaster Telephone Directory and for the number of letters in the items. We refer to these as unambiguous surnames. Finally, eight occupations which are not commonly used as surnames were chosen (architect, grocer, solicitor, etc.). These unambiguous occupations were matched to the ambiguous labels on word frequency using the Kucera & Francis (1967) norms, and on number of letters. The characteristics of these matched sets of items are given in Table 1. Table 1. Characteristics of matched sets of items used in the experiment Mean Mean Mean Telephone Kucera— number Directory ' Francis of Description frequency frequency letters Set A Ambiguous labels 1350 14-25 638 Set B Ambiguous labels 13-75 13-00 6- l 3 Set C Unambiguous surnames 13-75 — 6-00 Set D Unambiguous occupations — 13-25 6-88 These sets of items were arranged into two groups of stimuli. The first group used one of the sets of ambiguous labels as surnames coupled to the unambiguous occupations, and the unambiguous surnames with the other set of ambiguous labels as occupations. In the second group of stimuli use of the two sets of ambiguous labels as occupations or as surnames was reversed. Thus the same ambiguous label did not appear as both surname and occupation in the same group of stimuli, and each pairing of surname and occupation only contained one ambiguous label. Use of these two groups of stimuli was counterbalanced across subjects in order to ensure that as many subjects learnt each ambiguous label as a surname as learnt that label as an occupation. Thus subjects learnt occupations and surnames belonging to 16 faces, to a criterion of two correct runs through the entire 16 faces. The occupations that they learnt could be unambiguous (i.e. occupations that are not usually surnames) or ambiguous (occupations that could be surnames), and the surnames could also be unambiguous (i.e. surnames that are not occupations) or ambiguous (Surnames that could be occupations). Errors made while learning these occupations and surnames were recorded. Results On each experimental trial the subject was presented with a face and asked to give the PCrSon’s occupation and surname. On average, subjects needed to go through the sets of 16 faces 7-38 times (SD 141) in order to reach our criterion of two entirely correct runs through the 16 faces. The responses for each trial can effectively be divided into four types: those in which b0th occupation and surname were correctly given, those in which only the occupation was 6-2 146 Kathryn H. McWeeny, A. W. Young, D. C. Hay and A. W. Ellis correctly given, those in which only the surname was correctly given, and those in Which' neither item was correctly given. The mean number of trials on which both occupatjon and surname were correctly given was 66-06 (SD 11-49), and the mean number of trials on which neither item was correctly given was 2494 (SD 10-24). However, it is the numbers 0 trials in which only the occupation was correctly given and the numbers of trials in Which only the name was correctly given that are of most interest. The numbers of occasions on which only the occupation was correctly given or only the name was correctly given were totalled for each subject. These totals were then subdivided into those in which the unrecalled item belonged to one of the ambiguous or unambiguo“s sets. Means and standard deviations are presented in Table 2, where the data are also subdivided according to whether subjects were made to learn and recall the items in the order name then occupation or in the order occupation then name. Table 2. Means and standard deviations of numbers of trials on which subjects recalled only an occupation or only a name \- . Occupation recalled Name recalled Unrecalled Unrecalled Unrecalled Unrecalled Order of ambiguous unambiguous ambiguous unambiguous learning name name occupation occupation Name then occupation SD 1163 6-16 9-25 5-68 0-38 0-74 0-63 0-74 Occupation X— 15-25 12-13 1-13 1-13 then name SD 7-69 2-90 2-03 1-81 Overall 7 13-44 10-69 0-76 0-88 SD 6-99 4-60 1-53 136 The most striking feature of the data was that every subject recalled an occupation without a name more often than s/he recalled a name without an occupation; this observation held for both the ambiguous and the unambiguous occupations and names. In :. fact the occasions on which a name was remembered without an occupation were so rare that the data in these cells of the design showed little variance. An Fmax test applied to the . data as presented in Table 2, following the procedure recommended by Winer (1962), showed that the assumption of homogeneity of variance necessary for an analysis of variance had been violated (Fmax = 10727, k = 8, n = 7, P < 0-001). For this reason the data were subjected to a square root transformation. The effect of this transform was to reduce variances in the cells containing the larger scores, and thus satisfy the assumption of homogeneity of variance (Fmax = 290, k = 8, n = 7, P > 0-1). A three-factor analysis of variance of these transformed data was carried out to determine the effects of order of learning (name then occupation, or occupation then name), item recalled (occupation recalled or name recalled; repeated measure), and item ambiguity (ambiguous or unambiguous; repeated measure). This revealed only one statistically significant effect, in which occupations were recalled more often than names (item recalled, F: 139-88, d.f. = 1,14, P < 0-001). The results of the experiment thus show that surnames are harder to recall than occupations, regardless of the order in which the items are learnt and regardless of their ambiguity or lack of ambiguity. In orde particular groups 0f expel-fine] (binomial harder to Discussio The expe of [CtI’lCV occupatir related it semantic informal every tri occupati The I'( have her This fin: (name ti and for Mr Pot1 This : semanti Young nodes’ ' person that Sp: systems recogni Young takes p 1986) a (McKe The name i for err should Nonet . W. Ellis :tly given, and those in which : on which both occupation and : mean number of trials on ). However, it is the numbers of the numbers of trials in which was correctly given or only the ese totals were then subdivided :he ambiguous or unambiguous 2, where the data are also n and recall the items in the 1 name. Is on which subjects recalled 3 Name recalled alled Unrecalled uous unambiguous ition occupation 8 0-63 4 0-74 3 H3 3 1-81 6 0-88 3 1-36 ct recalled an occupation ut an occupation; this ous occupations and names. In it an occupation were so rare :e. An Fmax test applied to the mended by Winer (I962), essary for an analysis of < 0001). For this reason the ect of this transform was to 1 thus satisfy the assumption of .). ata was carried out to Ltion, or occupation then repeated measure), and item This revealed only one illed more often than names : harder to recall than :arnt and regardless of their '7 l l l Recalling names 147 In order to check that the effect was not simply due to a few surnames being particularly diflicult to recall, an items analysis was also carried out. The use of two groups of stimuli meant that a total of 32 name+occupation pairings were used in the experiment. For 31 of these pairings the name was the more frequently forgotten item (binomial test, 2 = 513, P < 0-001). Thus it is clear that the finding that surnames are harder to recall than occupations is of general applicability. Discussion The experiment was designed to eliminate factors that might underlie the relative difficulty of retrieving names as compared to other semantic information. Thus contextual cues to occupation were eliminated by presenting the faces in isolation, cueing from the retrieval of related items of semantic information was eliminated by only giving subjects one piece of semantic information (the person’s occupation), and frequency of use of semantic or name information was controlled by asking subjects to give the person’s name and occupation on every trial. In addition, ambiguous items were used as names for some subjects or as occupations for others. The results of the experiment are unequivocal. Even when all of the factors mentioned have been controlled, people’s names are much harder to recall than their occupations. This finding holds regardless of the order in which subjects learn and produce the items (name then occupation, or occupation then name), and it holds both for ambiguous and for unambiguous items. It is much more diflicult to recall that someone’s name is Mr Potter than to recall that he is a potter. This finding that people’s names are intrinsically harder to recall than other types of semantic information is predicted by the view taken by Young et al. (1985) and Bruce & Young (1986), who argue that as we learn about familiar people we create ‘person identity nodes’ which hold essential semantic codes and that names are then accessed via these person identity nodes. The pur...
View Full Document

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern