Gender-Linked Differences in Everyday Memory Performance: Effort Makes the Difference

Gender-linked differences in everyday memory performance: Effort makes the difference

Colley, Ann;Ball, Jane;Kirby, Nicola;Harvey, Rebecca;Vingelen, Ingrid.Sex Roles47.11(Dec 2002): 577.

Turn on hit highlighting for speaking browsers

· Other formats:

· Full text

Abstract (summary)

Translate

Previous research has found gender differences in everyday memory tasks such as remembering shopping lists or directions, and these findings can be attributed to increased motivation or memory skill resulting from the association of different tasks with masculine or feminine gender roles. To investigate the motivational explanation, the recall of an ambiguous shopping list, labeled as grocery or hardware, was examined under instructions that stated that either women or men are good at the task, or instructions that were neutral. The findings were not consistent with the skill explanation. There was a significant interaction between list label, participant gender, and instructions, which can be explained in terms of the increased or decreased motivation that might arise from particular combinations of label and instruction.

Show less

Full Text

· Translate

Headnote

KEY WORDS: gender; memory; motivation.

Historically, memory research using standard laboratory tasks failed to demonstrate consistent gender differences (Maccoby & Jacklin, 1974). More recently, however, studies of episodic, autobiographical, and everyday memory have shown more consistent gender differences. An advantage for women has been found in many episodic memory tasks (e.g., Herlitz, Nilsson, & Baeckman, 1997; Lindholm & Christianson, 1998), particularly verbal episodic memory tasks or those that allow verbalization (e.g., Herlitz, Airaksinen, & Nordstroem, 1999; Lewin, Wolgers, & Herlitz, 2001). Women's superior verbal ability has been used as a partial explanation for their better performance in such tasks, but in some studies it has been suggested that gender differences in stored knowledge may play a role. For example, in Lindholm and Christianson's study (Lindholm & Christianson, 1998), women outperformed men in eyewitness memory for a violent murder, and the investigators suggested that women's memory of such events may be assisted by more elaborate categories for encoding information about people. Differences between men's and women's gender role socialization may result in variations in the amount of detail contained in the internal representations that guide information seeking and retrieval, and hence influence memory for gender-related information.

Such differences are particularly relevant to autobiographical memory. Women's recall of autobiographical events, such as childhood memories, has been found to be superior to men's (Friedman & Pines, 1991). Buckner and Fivush (1998) have found a similar gender difference in children's memory; girls recalled specific experiences at greater length and in more detail than boys did. In addition, girls recalled more interpersonal experiences, and, in describing these, their accounts contained more references to emotion, self, and affiliative themes. Other studies have confirmed an advantage for women in memory for emotion-related events. Davis (1999) found that women had fuller and faster recall than men of childhood memories linked to emotional experiences, and Seidlitz and Diener (1998) found that women recalled more positive and more negative life experiences than men did. Such findings are strongly suggestive of more efficient encoding of person- and emotion-related information by girls, perhaps as a result of their socialization into a person-oriented and communal gender role. Gender differences are not only present in the performance of tasks that measure the recall of personal memories, but also inform our expectations of memory performance. In a study of metamemory beliefs concerning men's and women's abilities to remember everyday information, Crawford, Herrmann, Holdsworth, Randall, and Robbins (1989) found that men's and women's beliefs about their own memory abilities differ and are consistent with their beliefs about the memory abilities of men and women in general. In their study, men and women were presented with a questionnaire adapted from Herrmann and Neisser's Inventory of Memory Experiences (Herrmann & Neisser, 1978). The questionnaire asked them to estimate how frequently they forgot information in everyday memory, for example, directions to a place only visited once before; the location of something put away a week earlier; items required from a supermarket but not on a written list. The participants were also asked to estimate the frequency of forgetting, both for men and for women in general. The findings indicated that women were believed to be better than men at remembering names, shopping lists, and childhood events, and at placing familiar names, whereas men were believed to be better than women at remembering directions. Herrmann, Crawford, and Holdsworth (1992) investigated whether the different gender-related expectations found by Crawford et al. were reflected in everyday memory performance. In the first of two experiments, they compared men's and women's recall of a shopping list (expected to be recalled better by women) and travel directions (expected to be recalled better by men). The anticipated gender differences were present in the recall data, which confirmed the expectations about differential performance. In the second experiment, the possibility that such gender differences may be moderated by the gender-linked content of the tasks was examined by using versions of the same tasks (shopping lists or directions) with male- or female-oriented labels. Thus a set of ambiguous directions was produced as either directions for making a workbench or a shirt, and a shopping list with appropriately ambiguous items was labeled as a grocery or hardware list. For both tasks men's and women's immediate recall varied in the predicted direction depending on the label used, although the effect was stronger for the directions task. A similar result was also obtained from the delayed recall of the directions task. Two possible explanations of these performance differences were proposed. Herrmann et al. argued that gender role demands create different opportunities for men and women to perform certain memory tasks. On tasks associated with masculine or feminine gender roles, metamemory beliefs could enhance performance by increasing effort for own-gender-typed tasks (motivation explanation). An alternative or additional explanation is similar to that outlined earlier in the discussion of findings from autobiographical and episodic memory tasks: differential memory skill that arises from role-related experience may lead to more efficient semantic processing (skill explanation). Although the motivation explanation might appear to be the most plausible explanation for the labeling effects found by Herrmann et al., the skill explanation cannot be ruled out because the gender cue may invoke relevant stored knowledge that facilitates encoding and that varies according to the different role-related experience of men and women. There is substantial evidence of such cueing effects in the memory literature (e.g., Bransford & Johnson, 1972). Findings from other types of memory tasks can also be explained by either or both of the motivation and skill explanations; for example, Cherney and Ryalls (1999) found that both children and adults had better incidental recall for same-gender-congruent objects together with their location in a room previously visited. Again, disposition to notice and encode samegender-congruent objects may have resulted in differential encoding, or such objects may be more readily encoded in the context of existing knowledge. The use of existing knowledge to facilitate encoding has been demonstrated in a number of studies of skilled memory, including Chase and Ericsson's observations of SF (Chase & Ericsson, 1982), a long distance runner who increased his digit span from around 7 to 82 by using, among other encoding strategies, the association of three-digit chunks to running times for different distances. SF's performance was enhanced by both the effort exerted and the use of prior specialist knowledge. Other findings in the memory literature provide a clearer demonstration of the role of motivation or disposition to respond by varying gender cues but not the material to be memorized. For example, Crawford and English (1994) presented men and women with essays in which the target was described as "he" or "he or she" and found that men recalled more content when "he" was used, whereas women recalled more in the "he or she" condition. In this case the variation between conditions lies simply in the pronouns used and does not map directly on to the content of the memory task or the existing knowledge of the participants. Evidence of the effect of gender cues in producing a motivational effect is not confined to memory tasks; gender-typed labeling has also been found to produce differential performance by boys and girls in perceptual motor games (Montemayor, 1974) and by men and women in laboratory tasks (e.g., Hargreaves, Bates, & Foot, 1985). However, perhaps the most striking findings of differential performance in the presence of gender-related cues come from the literature on stereotype threat (Steele & Aronson, 1995). Stereotype threat arises from the association of a lower level of ability in a particular domain with an identifiable social group. When such group identity is made salient, apprehension arises from fear of confirming a negative stereotype and results in underperformance by members of the stigmatized group. Steele and Aronson (1995) demonstrated that African Americans performed much worse on a difficult verbal test when told that it was diagnostic of intelligence than when told that it was not. They also performed worse under nondiagnostic conditions after having their racial identity primed. No such differences were present for European Americans. Similarly women underperformed relative to men on difficult math tasks, and this difference increased when a description indicated that the test produced gender differences and disappeared when a description indicated that it did not produce gender differences (Spencer, Steele, & Quinn, 1999) The present experiment was designed to examine gender-related motivational effects in an everyday memory task by not only providing a gender cue in the label applied to the information presented, but also by manipulating information provided about men's or women's performance. Using Hermann et al.'s list recall task (Hermann et al., 1992), a shopping list labeled as either "grocery" or "hardware" was presented with instructions that indicated better performance by men or better performance by women. In addition, a neutral instructions condition was used. If being told that men or women are good at the task results in recall performance that is different from that obtained without such information, then it would seem reasonable to conclude that motivation can play a role in list recall. From the existing literature it was predicted that that poorer performance would result when the list was given a label associated with better performance by the other sex and that this effect would be strengthened when participants were told that the other sex is good at the task. METHOD Participants Sixty male and 60 female British undergraduates (White, mean age = 20.5 years, SD = 3.1) opted to participate in the study in partial fulfillment of a course requirement. Design The participants were allocated to one of six experimental groups. Each group memorized a list headed either "Grocery Shopping List" or "Hardware Shopping List" with one of three sets of instructions that varied in their opening sentence: "This is a study of the kind of everyday memory we all use frequently"; "This is a study of the kind of everyday memory men are good at"; "This is a study of the kind of everyday memory women are good at." Ten men and 10 women took part in each combination of list and instructions. In common with the study of Hermann et al. (1992) both immediate and delayed recall conditions were used. Materials The shopping list consisted of 16 items selected to be plausible constituents for both a grocery and a hardware shopping list. The list differed somewhat from those used by Herrmann et al. (1992) in order to be appropriate for a British sample: detergent, glue, candles, oil, dye, nuts, salt, polish, matches, disinfectant, bleach, scourer, charcoal, soap, wax, and clingfilm. Procedure After reading a written description of the task they would be required to undertake, the participants were given 45 s to memorize as many items as possible from the written list, then they wrote down as many items as they could remember. Their response sheet was then removed, and they were given a 5-min distracter task in which they read a lengthy sheet of text (taken from "The Royal Charter the Statutes the Rules" of the British Psychological Society, 1994) and crossed out as many consonants as was possible in the time allotted. At the end of the 5-min period they again wrote down as many items as they could recall from the original list.

RESULTS

In order to examine the effect of the list label alone and compare our results with those of Herrmann et al., a 2 x 2 x 2 (Participant sex x List x Recall condition) mixed ANOVA of data from the groups who received neutral instructions was undertaken, followed by analysis of the whole design.

Neutral Instructions

It is not surprising that recall was significantly better in the immediate than the delayed condition (immediate recall mean = 9.58, SD = 2.58; delayed recall mean = 8.75, SD = 2.59), F(1, 36) = 8.68, p = .006. More interesting is the finding that the recall of items from the hardware list by men was similar to that of women, but the recall of items from the grocery list was better by women than by men, F(1, 36) = 4.67, p = .04. The means from this interaction, which was based upon combined data from the two recall conditions, are shown in Fig. 1. The two recall conditions each produced a very similar pattern of results.

Overall Analysis