- Email: [email protected]
Mental rotation performance in primary school age children: Are there gender differences in chronometric tests?
Cognitive Development 28 (2013) 51–62
Contents lists available at SciVerse ScienceDirect
Cognitive Development
Mental rotation performance in primary school age children: Are there gender differences in chronometric tests? P. Jansen a,∗, A. Schmelter a, C. Quaiser-Pohl b, S. Neuburger b, M. Heil c a b c
Universität Regensburg, Institute of Sport Science, Regensburg, Germany Universität Koblenz, Institute of Psychology, Koblenz, Germany Heinrich-Heine Universität Düsseldorf, Institute of Experimental Psychology, Düsseldorf, Germany
a r t i c l e Keywords: Mental rotation Pre-adolescence Gender differences Stimulus type
i n f o
a b s t r a c t In contrast to the well documented male advantage in psychometric mental rotation tests, gender differences in chronometric experimental designs are still under dispute. Therefore, a systematic investigation of gender differences in mental rotation performance in primary-school children is presented in this paper. A chronometric mental rotation task was used to test 449 second and fourth graders. The children were tested in three separate groups each with different stimulus material (animal drawings, letters, or cube figures). The results show that chronometric mental rotation tasks with cube figures – even rotated in picture plane only – were too difficult for children in both age groups. Further analyses with animal drawings and letters as stimuli revealed an overall gender difference in response time (RT) favoring males, an increasing RT with increasing angular disparity for all children, and faster RTs for fourth graders compared to second graders. This is the first study which has shown consistent gender differences in chronometric mental rotation with primary school aged children regarding reaction time and accuracy while considering appropriate stimuli. © 2012 Elsevier Inc. All rights reserved.
∗ Corresponding author at: Institute of Sport Science, Universitätsstr. 31, 93053 Regensburg, Germany. Tel.: +49 941 943 2518; fax: +49 941 943 4490. E-mail address: [email protected] (P. Jansen). 0885-2014/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cogdev.2012.08.005
52
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
1. Introduction Mental rotation is the ability to rotate two- or three-dimensional stimuli in the mind (Shepard & Metzler, 1971). The main goal of this study is to investigate gender differences in the chronometric mental rotation performance of school age children when using different types of stimuli. This is a combination of topics which has not been systematically investigated until now. 1.1. Mental rotation performance measured with psychometric and chronometric tests in children Psychometric mental rotation tests, where participants have to solve time-limited mental rotation tasks on a paper sheet, can be used to test children as young as 5 years old (Quaiser-Pohl, Rohe, & Amberger, 2010) when using age appropriate stimuli (Quaiser-Pohl, 2003). With the most widely used paper–pencil mental rotation test, the MRT (Peters et al., 1995; Vandenberg & Kuse, 1978), psychometric mental rotation performance can be reliably measured in primary school-aged children (9–11 years) (Titze, Jansen, & Heil, 2010). In the MRT, each of the 24 items consists of one target on the left side and four comparison items on the right side. Two of the four comparisons are “correct” (identical to the left item) and two are “incorrect”. Participants must choose both of the correct stimuli within a time limit of three minutes for 12 items. Accuracy is measured. In addition to psychometric tests there are chronometric mental rotation tests. In these tests, participants sit in front of a computer screen and are shown pairs of mental rotation objects. They must decide as fast and as accurately as possible if the objects are the same or are mirror images of each other by pressing a key. Accuracy and reaction time are measured. It has been demonstrated that children are able to solve chronometric mental rotation tasks even before entering school. Marmor (1975) showed that 5-year-old children can mentally rotate two-dimensional figures in the picture plane. However, they were twice as slow as 8-year-old children. These results were confirmed in other studies (e.g. Kosslyn, Margolis, Barrett, & Goldknopf, 1990), but they are also still controversially discussed (Newcombe, 2002). Results from additional studies have revealed a high individual variability and some studies have failed to replicate Marmor’s result (compare Newcombe & Frick, 2010). Kail, Pellegrino, and Carter (1980) investigated the development of mental rotation performance after the age of eight. They found a linear mental-rotation function using alphanumeric and abstract symbols and showed that mental-rotation speed nearly doubled from 8 years old to adulthood. Concerning the influence of stimulus type, Courbois (2000) showed that the ability to rotate unfamiliar stimuli with no salient axis improves from five to 8 years of age. Furthermore, Perrucci, Agnoli, and Albiero (2008) demonstrated that children at the age of 6 years old are already able to compare two stimuli at different orientations dependent on orientation-free features, such as the color of some part of the stimuli. In this case they are able to respond without using mental rotation. 1.2. The importance of gender differences in children measured with psychometric and chronometric tests Gender differences in mental-rotation tasks are widely discussed and well investigated in psychometric, paper–pencil tests (Peters & Battista, 2008). Using psychometric tests with three different stimulus conditions (animal pictures, letters, and cube figures) it was shown that fourth graders, but not second graders, showed a small, significant, stimulus-independent gender difference favoring males (Neuburger, Jansen, Heil, & Quaiser-Pohl, 2011). This is in line with another study showing a large gender difference favoring males only in older (mean age: 10.3 years) but not in younger fourth graders (mean age: 9.3 years) (Titze et al., 2010). These gender differences are often discussed on a psycho-social or biological–neuronal basis. Psycho-social theories often argue that gender differences are based on the influence of attitudes or stereotypes (Moè & Pazzaglia, 2006; Steele, 1997) or on different experiences based on gender. For example, Levine, Ratliff, Huttenlocher, and Cannon 2012) showed that the quality, but not the frequency, of puzzle play was higher for boys than for girls (between 2 and 4 years old) and that this variation predicts performance on spatial transformation tasks. Biological theories stress the importance of brain organization and hormonal (organizational and activational) influences. Both psycho-social and biological approaches are empirically supported
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
53
and their factors seem to interact in a complex way (Casey, Colon, & Goris, 1992). Furthermore, especially with children, task attributes in the mental rotation task itself also play an important role in mental rotation performance (e.g. Courbois, 2000; Perrucci et al., 2008). Gender differences in chronometric mental-rotation tasks have been investigated more thoroughly in adulthood than in childhood. Seven (5 with children under 14 years old) of the 15 chronometric studies in the meta-analysis of Voyer, Voyer, and Bryden (1995) showed no gender differences at all, resulting in a small to medium overall effect size (d = .37). However, studies are difficult to compare because they vary with respect to stimulus types, sample sizes, and the analyzed variables of information processing. In a systematic study with a total of N = 360 adult participants, Jansen-Osmann and Heil (2007a) demonstrated that only polygons, but not characters, animal drawings, PMA symbols (2D geometric objects from the Primary Mental Ability Battery, Thurstone, 1958), or cube figures, produced a substantial and reliable gender difference in mental-rotation speed. This suggests that the finding of a male advantage in psychometric tests cannot automatically be generalized to chronometric mentalrotation tasks. Studies concerning gender differences with children using chronometric tests are rare. Krüger and Krist (2009) used pictures of body parts as stimuli material and found an unexpected gender difference: kindergartner’s boys made more errors than girls. In another study, Heil and Jansen (2010) investigated gender differences in both psychometric and chronometric mental rotation tests and in a standardized math test for 7–8-year-old children (47 girls, and 62 boys). Gender differences favoring boys were only found in accuracy measurements (psychometric mental rotation test, math test, and error rate of the chronometric test). 1.3. Goal of this study Regarding the use of chronometric mental rotation tests, a systematic investigation regarding gender difference, as it was presented for adults (Jansen-Osmann & Heil, 2007b), is missing for school age children. The main goal of this study is to investigate the gender differences present in school age children using a chronometric test and to assess how appropriate different stimuli are for this age group. According to the study of Heil and Jansen (2010) a gender difference in a chronometric mental rotation test using animal drawings is anticipated for accuracy measurements. The study presented here will add to their findings because we are additionally investigating the influence of different types of stimuli and different age groups on chronometric mental rotation performance. The study presented here will add to their findings because we are additionally investigating the influence of different types of stimuli and different age groups on chronometric mental rotation performance. Additionally, general intelligence and socioeconomic status were controlled, because socioeconomic status has been shown to modify gender differences in spatial tasks (Levine, Vasilyeva, Lourenco, Newcombe, & Huttenlocher, 2005). Because cube figures are more difficult to process than animal drawings or letters, indicated by longer reaction times (e.g. Jansen-Osmann & Heil, 2007a), we expected a larger gender difference for the cube figures. Furthermore, we expected a better mental rotation performance from fourth graders compared to second graders due to an increase in general processing speed (e.g. Kail, 1993) and rotation rate (Kail et al., 1980) with age. 2. Method 2.1. Participants The participants in this study consisted of 449 elementary-school children (228 second graders, 221 fourth graders) from schools in the area of Bonn, Western Germany. All parents gave their written, informed consent. The sample included children from families with low (13.4%), middle (14.3%), high (17.2%), and very high (48.7%) socio-economic status (SES). 6.4% did not complete the SES questionnaire. The socioeconomic status was measured by a questionnaire based on the measure provided by Jöckel et al. (1998). In this questionnaire, parents of the participants were asked to indicate their school leaving certificate and professional degree. Following Jöckel et al. (1998), rank-ordered values were assigned to the different combinations of graduation levels.
54
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
To ensure that we only used data from children that did attempt to solve the mental rotation test as instructed, we excluded the data of children whose error rates in at least one of the experimental conditions exceeded 90%. This resulted in the exclusion of 19 children, 12 boys (9 second graders) and 7 girls (6 second graders. From the remaining 430 children, 212 children were second graders (mean age: 7.94 years, SD = .5, range: 6.83–9.92 years), and 218 were fourth graders (mean age: 10.06, years, SD = .54, range: 8.25–12.17 years). Each child received little presents (value: .5D ) for participation. Additionally, participating classes were given 4D per child for the class treasury. Non-parametric statistics (Mann–Whitney U-test) showed that the SES did not differ between boys and girls (p = .53), second and fourth graders (p = .67), or the three stimulus condition groups (p = .67). Because of this, we could exclude that possible gender differences are modified by the socioeconomic status. Each child was tested in one of the three stimulus types (animal drawings, letters, or cube figures). The following numbers of children were analyzed in the different stimulus types: “animal drawings”: 60 second graders (32 girls and 28 boys) and 67 fourth graders (31 girls and 36 boys); “letters”: 85 second graders (48 girls and 37 boys) and 77 fourth graders (39 boys and 38 girls); “cube figures”: 67 second graders (34 girls and 33 boys) and 74 fourth graders (37 boys and 37 girls). 2.2. Material The three stimulus conditions were completely equivalent besides the stimuli used. The “animal drawings” condition consisted of colored drawings of 20 different animals (bear, camel, raccoon, cow, crocodile, dog, donkey, elephant, goat, grizzly bear, horse, leopard, lion, monkey, pig, rhino, sheep, tiger, turtle, and zebra, respectively, from Snodgrass and Vanderwart (1980)). The “letters” condition consisted of 15 alphanumeric letters (e, F, g (twice but different fonts), j, k (twice but different fonts), L, l, n (twice but different fonts), P, q (three times but different fonts) R, r, s, S, z). In condition 3, we used “cube figures” similar to those by Shepard and Metzler (1971). An example of each stimulus type is given in Fig. 1. In order to parallelize task demands for all three stimuli types and to avoid floor effects in secondgrade participants, we only used picture-plane rotations for the cube figures. Each stimulus had a maximal size of 7 by 7 cm on the screen with a space of 14 cm in between. The children were free to choose the most comfortable viewing distance. In each stimulus condition, the right stimulus was either identical to the left or mirror-reversed. The left stimulus was always presented upright. The angular disparity between the two stimuli was 45◦ , 90◦ , or 135◦ (clockwise or counter clockwise). With “animal drawings” and “letters” as stimuli each child had to complete 240 trials each. Because mental rotation with cube figures as the stimulus material was supposed to be much more difficult, each child only had to complete 120 trials to avoid overstraining the participants. After 20 trials a short break was administered in all stimuli conditions. The experiment was run on a laptop with a 15 monitor and lasted approximately 1 h. General cognitive abilities were controlled by administering the subtest “Reasoning” of the “Cognitive-Ability Test” (KFT 1–3, Heller & Geisler, 1983). This test consists of 15 items, which are made up of five pictures, respectively. Four of the five pictures belong to the same category (e.g. vegetable). The picture that does not fit into this category has to be detected and marked by the child. There is no time constraint. The test was chosen because it measures general non-verbal cognitive abilities, and because performance in this test does not depend on visual-spatial abilities. Furthermore, it can be administered in groups. 2.3. Procedure Children were tested in a quiet room during regular school time in groups of two to five children. The children were separated in this room so that every child was able to solve the tasks individually on a laptop. After a short introduction, children were given the KFT subtest. Since the KFT is a power test, there was no temporal constraint. When all children had finished with the KFT, the mental rotation task started. Each session was preceded by 40 unrecorded practice trials. Children were allowed to choose their own pace by pressing a key when they wanted to start the next run of 20 trials. Each trial started with the presentation of a 500 ms background gray screen. After this, the stimuli pair
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
55
Fig. 1. Sample items of the mental rotations tasks (animal drawings, letters, and cube figures).
was presented. Children had to decide as fast as possible if the stimuli were the “same” or “different” (mirror reversed). They were instructed to press the green marked mouse button when the stimuli were the “same” and the red marked mouse button when the stimuli were “different”. For pressing the key they were allowed to choose which finger to respond with. The children received feedback in the form of a “+” for correct responses and a “−” for incorrect responses only in the 40 training trials. Both feedbacks appeared for 500 ms in the center of the screen. The next trial began after 1500 ms. No feedback was given in the non-training trials. 3. Results For the statistical analysis of response time (RT), only trials with correct responses were used. RTs more than 2 SDs above or below the mean per condition and per participant were excluded, resulting in the exclusion of a total of 6.2% of the data. Analyses with untrimmed data revealed identical results. For all analyses clockwise and counter clockwise angular disparities were averaged. 3.1. Reasoning score Reasoning ability was defined as the number of correctly answered items in the KFT subtest and transformed into age-corrected standard values. An ANOVA with the KFT-score as a dependent variable
56
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
and the between subject factors “grade” (grade 2 vs. grade 4), “gender” (boys vs. girls), and “stimulus type” (animal drawings vs. letters vs. cube figures) was conducted. There were significant effects of the factors “gender”, F(1,418) = 13.60, p < .001, 2 = .03, and “grade”, F(1,418) = 35.02, p < .001, 2 = .08, but no significant main effect of the factor “stimulus type”, (F(2,418) = 2.05, n.s. The children in the 2nd grade (M = 47.34, SE = .64) had a lower score than the children in the 4th grade (M = 52.68, SE = .63), and boys (M = 48.34, SE = .64) had a lower score than girls (M = 51.67, SE = .63). The interaction between both factors, F(1,418) = 3.76, p = .053, 2 = .01, did not reach significance. In the 2nd grade, girls (M = 49.29, SE = .90) had a higher score than boys (M = 44.85, SE = .99), F(1,210) = 13.50, p < .001, 2 = .06. The score did not differ between girls (M = 53.59, SE = .87) and boys (M = 51.91, SE = .81) in the 4th grade, F(1,216) = 2.01, n.s. The KFT-score was considered as a covariate in the following analyses. 3.2. Accuracy score An error measurement was calculated for each participant. In reference to memory research (Snodgrass & Corwin, 1988), the PR-score (which is the abbreviation for the discrimination index according to Snodgrass and Corwin (1988)) was calculated for each angular disparity (45◦ , 90◦ , 135◦ ). The PR score is defined as the difference between the hits percentage (percentage of “same” responses for trials where “same” was the correct response) and the false alarm percentage (percentage of “same” responses for trials where “same” was the incorrect response). This measurement allows for the correction of guessing by chance and of guessing by always pressing the same button. An ANCOVA was conducted covarying the KFT score with the dependent measure “PR-score”, the within subject factor “angular disparity” (45◦ , 90◦ , 135◦ ), and the between subject factors “stimulus type” (animal drawings vs. letters vs. cube figures), “gender” (boys vs. girls), and “grade” (grade 2 vs. grade 4). The significance levels were corrected according to Huynh and Feldt (1976) in order to compensate for the non-sphericity of the data. The ANCOVA revealed a significant main effect for four factors: “angular disparity“, F(2, 834) = 5.28, p < .01, 2 = .01; “grade”, F(1,417) = 14.24, p < .001, 2 = .03; “gender”, F(1,417) = 11.52, p < .01, 2 = .03; and “stimulus type”, F(2,417) = 150.27, p < .001, 2 = .42. Furthermore, there was a significant interaction between the factors “grade” and “stimulus type”, F(2,417) = 7.37, p < .01, 2 = .03, between the factors “angular disparity” and “stimulus type”, F(4,834) = 3.60, p < .05, 2 = .02, and a three-way interaction between “angular disparity”, “stimulus type”, and “gender”, F(4,834) = 3.63, p < .05, 2 = .02 (see Fig. 2a and b).1 In Fig. 2a the accuracy score for the stimulus type “animal drawings” is presented showing a significant main effect of the factor “gender”, F(1,124) = 5.83, p < .05, 2 = .05, but not for “angular disparity”, F(1,124) = 3.00, n.s. However it does show a significant interaction between both factors, F(2,248) = 5.46, p < .05, 2 = .04. There was neither a gender difference for an angular disparity of 45◦ , F(1,125) = .13, n.s., nor for an angular disparity of 90◦ , F(1,125) = 2.58, n.s., but there was a gender difference for an angular disparity of 135◦ , F(1,125) = 5.32, p < .05, 2 = .04. Fig. 2b shows the accuracy score for the stimulus type “letters”. There was a significant main effect for “gender”, F(1,159) = 5.18, p < .05, 2 = .03 and “angular disparity”, F(2,318) = 8.16, p < .001, 2 = .05, 1 Because the PR-score is rarely used in the analysis of the accuracy rate in mental rotation the results were compared to the often used measurements of accuracy rate in mental rotation: analyzing only the accuracy rate for the responses on items which were the same showed a result which was comparable with the PR-score measurement regarding the main effects. The ANCOVA for the accuracy rate on same items revealed a significant main effect for four factors: “angular disparity”, F(2,834) = 13.82, p < .001, 2 = .03; “grade”, F(1,417) = 9.65, p < .01, 2 = .02; “gender” F(1,417) = 5.78, p < .05, 2 = .02; and “stimulus type”, F(2,417) = 33.96, p < .001, 2 = .14. Furthermore, there was a significant interaction between the factors “grade” and “stimulus type”, F(2,417) = 4.43, p < .01, 2 = .02, and a three-way interaction between “angular disparity”, “stimulus type”, and “grade”, F(4,834) = 2.55, p < .05, 2 = .01. When analyzing the accuracy rate for the responses on items which were different, the ANCOVA revealed a result which was comparable of the ANCOVA of the PR-score. A main effect of angular disparity was missing despite the significant interaction between angular disparity and KFT-score, F(2, 834) = 3.23, p < .05, 2 = .008. Furthermore, there were significant main effects of “grade”, F(1,417) = 9.18, p < .01, 2 = .02; “gender” F(1,417) = 9.39, p < .05, 2 = .02; and “stimulus type”, F(2,417) = 182.89, p < .001, 2 = .47. Additionally there were interactions between “grade” and “stimulus type”, F(2,417) = 5.39, p < .01, 2 = .03 and between “angular disparity”, “stimulus type”, and “gender”, F(4,834) = 2.91, p < .05, 2 = .01.
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
57
Fig. 2. PR-score (means and standard deviation) dependent on stimulus type, angular disparity, and gender. Part (a) shows the PR-score for animal drawings; part (b) shows the PR-scores for letters.
but no interaction between both factors, F(2,318) = 1.3, n.s. Boys (74.53%, SE = 2.93) had a greater accuracy score than girls (70.35%, SE = 3.00). The accuracy score was higher for an angular disparity of 45 ◦ (M = 80.07%, SE = 2.03), compared to 90 ◦ (M = 73.22%, SE = 2.24), F(1,161) = 46.46, p < .001, 2 = .23, compared to 135◦ (M = 63.64%, SE = 2.33), F(1,161) = 77.61, p < .001, 2 = .33. Concerning the accuracy score for the stimulus type “cube figures” there was no significant main effect of “angular disparity”, F(2,276) = .34, n.s., or “gender”, F(1,138) = 2.03, n.s., nor a significant interaction between both factors, F(2,276) = .76, n.s. The descriptive effect of angular disparity did not reach significance because the KFT-score was used as a covariate, F(2,276) = 3.84, p < .05, 2 = .03. The
58
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
Fig. 3. Reaction time (means and standard deviation) dependent on stimulus type and grade.
performance in the cube condition (averaged over all angular disparities) was beneath chance for girls (M = 37.78%, SE = 3.46) as well as for boys (M = 35.56%, SE = 3.43). 3.3. Response time Response time data of two children were excluded because their reaction time was more than three standard deviations above the mean. Because of the very poor PR-score for the stimulus type “cube figures” (see Section 3.2) further analyses concerning response time were only conducted with the stimuli “animal drawings” and “letters”. For these two stimulus types, all children who had an error rate lower than 30% in one of the six conditions (3 levels of angular disparity: 45◦ , 90◦ , 135◦ ; 2 types of stimuli presentation: identical or mirror-reversed) were included in further analysis (43 girls and 43 boys in grade 2, and 53 girls and 64 boys in grade 4). The cut-off criterion was decided based on the work of Perrucci et al. (2008), who only allowed children to participate in a mental rotation task when they had solved a pre-test stimulus matching task with less than 18% errors, and on the work of Wiedenbauer and Jansen-Osmann (2008) who excluded all children with an error rate over 40%. The KFT reasoning score was used as a covariate in the analysis of the response time. This analysis of covariance with response time as the dependent variable was restricted to responses to identical items only, because angular disparity is not unequivocally defined for responses to mirror-reversed items. The within subject factor was defined as “angular disparity” (45◦ , 90◦ , 135◦ ), and the between subject factors were defined as “gender” (boys vs. girls), “grade” (grade 2 vs. grade 4), and “stimulus type” (letters vs. animal drawings). The significance levels were corrected according to Huynh and Feldt (1976) in order to compensate for the non-sphericity of the data. Reliability of the data was given by measuring the split-half reliability using the odd-even method and the Spearman–Brown prediction formula for each grade and each angular disparity (grade 2: 45◦ : r = .72, 90◦ : r = .8, 135◦ : r = .83; grade 4: 45◦ : r = .77, 90◦ : r = .51, 135◦ : r = .86). The ANCOVA showed significant main effects of the factors “gender”, F(1,193) = 10.36, p < .005, 2 = .05, “grade”, F(1,193) = 32.98, p < .001, 2 = .15, and “angular disparity”, F(2,386) = 39.65, p < .001, 2 = .17, but not for “stimulus type”, F(1,193) = .81, n.s. (see Fig. 3). There was no significant interaction. Girls (M = 2000.30 ms, SE = 52.94) needed more time to solve the task than boys (M = 1817.68, SE = 45.98). Children in the 2nd grade (M = 2147.26 ms, SE = 55.19) showed longer response times than
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
59
children in 4th grade (M = 1720.79 ms, SE = 38.80), and furthermore, the response time increased with increasing angular disparity (45◦ : M = 1505.68 ms, SE = 30.99; 90◦ : M = 1839.72 ms, SE = 36.14; 135◦ : M = 2368.01 ms, SE = 47.00). 4. Discussion This is the first study which investigates the influence of grade, gender, and stimulus type, and their possible interactions, on the chronometric mental-rotation performance of elementary-school children. Previously these elements have only been explored in psychometric tests. Our first hypothesis regarding gender differences could only be partially confirmed: with animal drawings boys had a higher accuracy rate only at 135◦ of rotation; with letters the overall better accuracy rate of boys could be shown at all angles. Secondly, it was shown that boys were faster than girls at all rotational angles, and fourth graders were faster than second graders. Furthermore, the results show that most children in the 2nd grade and the 4th grade had substantial difficulties to solve a chronometric mental rotation test with cube figures rotated in the picture plane. Because of this, our hypothesis that cube figures lead to a larger gender difference could not be investigated in detail. 4.1. Gender differences in accuracy score Concerning accuracy rate, the results are different for letters and animals drawings. With animal drawings boys had a higher accuracy rate only at 135◦ of rotation; with letters an overall better accuracy rate of boys was demonstrated. The gender difference in accuracy score with animal drawings only at the higher angular disparity of 135◦ could be explained if the boys used a different strategy than the girls to solve these trials. Error rates are more sensitive than reaction time to the effect of holistic vs. piecemeal rotation. According to different sources of error in the matching process accuracy degrades as a function of angular disparity when adopting a piecemeal strategy, where participants compare each part of the figure with the corresponding parts of the reference object (Amorim, Isableu, & Jarraya, 2006). The higher error rate for girls with animal drawings at 135◦ suggests that more girls may use a piecemeal strategy while rotating animal drawings. Another explanation might be that the animal drawings are “embodied”. This strategy involves mapping the body axes (head–feet, front–back, and left–right) onto the animal drawing. If this strategy was used it might be assumed that girls had more difficulties to rotate the picture into the unfamiliar body positions, such as 135◦ compared to 45◦ and 90◦ . This could be due to the girls having less motor experience with these angles than the boys, but this remains rather speculative and has to be investigated further. Boys showed an advantage in accuracy score for letters compared to girls at all angular disparities. This might indicate that the strategies between boys and girls did not differ at either age while solving a mental rotation task with letters. Differentiating which strategy is used between boys and girls could be investigated with eye-tracking analysis (Just & Carpenter, 1985). 4.2. Gender differences in reaction time Concerning the reaction time when using letters and animal drawings as stimulus material, we found an increasing reaction time with increasing angular disparity, a longer reaction time for 2nd graders than for 4th graders, and a longer reaction time for girls than for boys. The decreasing reaction time from grade 2 to grade 4 is in accordance with a general increase in the speed of cognitive processes with age (Kail, 1988). The increasing reaction time with increasing angular disparity is in line with many other studies including the classic Shepard and Metzler (1971) study. Interestingly, this increase, which is assumed to reflect the process of mentally rotating the image, did not differ between boys and girls. Therefore, the effect of gender might be due to perceptual or motor processes executed before and after the rotation itself. In further studies a 0◦ condition has to be included to differentiate between the mental rotation process and the perceptual and motor processes in the mental rotation task. The obtained gender differences in this study are in accordance with Linn and Peterson (1985) who assumed that there is an emergence of gender difference favoring males as early as children can
60
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
mentally rotate. However, the meta-analysis of Linn and Peterson (1985) only included studies in which the psychometric MRT was used. Therefore, the revealed gender differences in chronometric mental-rotation tasks in primary school aged children are a new result. With regard to the influence of gender, these results are in contrast to the results of chronometric studies which used the same chronometric stimulus material with adults (Jansen-Osmann & Heil, 2007a). In their study no gender differences could be detected with letters and animal drawings as stimuli types. One possible explanation is that the different designs might contribute to these conflicting results: In the study with adults, trials were presented in blocks “always upright” or “sometimes upright” with angular disparities of 0◦ , 90◦ , and 180◦ , whereas in this study they were presented randomized and with angular disparities of 45◦ , 90◦ , and 135◦ . Another difference is that the participants in the adult study received feedback after each trial whereas the children only received feedback during the practice trials. While the women were able to see their “good” performance the girls did not.
4.3. Cube figures as stimuli A new result of the present study is that children were not able to solve the chronometric mentalrotation task with cube figures. To our knowledge there is no study investigating mental rotation in primary school-age children with cube figures as stimuli. For example, Kail et al. (1980) used alphanumeric characters and abstract symbols, Marmor (1975) used drawings rotated in picture plane, and Jansen-Osmann and Heil (2007b) used letters to investigate the neuronal correlates of the mental rotation process in primary school-aged children. Results are in accordance with the study of JansenOsmann and Heil (2007a) who showed that the mental rotation speed in adults was slower for cube figures than for letters, animal drawings, PMA-figures, and polygons. This means that adults also found it more difficult to solve a mental rotation task with cube figures than with other stimulus material. Children are less familiar with cube figures than with letters and animal drawings, this could cause the cube figures to be more difficult to process. These results suggest that the mental rotation performance in general is stimuli dependent and the stimuli used must be considered as a relevant factor in developmental and differential visual spatial research.
5. Conclusions On the basis of the present results, we can conclude that gender differences in chronometric mentalrotation tests favoring males deserve attention when letters and animal drawings are used as stimuli and that cube figures are too difficult for children between the ages of 8–10 years to process. This is a new finding because in previous chronometric studies gender differences are random (e.g. Voyer et al., 1995) or do only appear in accuracy measurements (Heil & Jansen, 2010). Finally, we know that our data will contribute to the discussion of gender differences in chronometric mental-rotation tests and reveal that experimental designs must be carefully considered when comparing studies of gender differences in mental rotation performance at this age. This study focused on children in middle childhood, but it is important to say that gender differences in spatial transformation task are found much earlier in life as demonstrated in preschool children (Levine, Huttenlocher, Taylor, & Langrock, 1999) and infants (Moore & Johnson, 2008; Quinn & Liben, 2008). According to our results the obtained gender differences in children younger than 8 years old might be also investigated regarding the different designs used. With this study we could not differentiate between a psycho-social and a biological–neuronal hypothesis. One concern is that “hormone levels begin to rise in middle childhood” (Archibald, Graber, & Brooks-Gunn, 2006, p. 25), and another concern is that, stereotypic conceptions of gender develop in middle childhood (Newcombe, Bandura, & Taylor, 1983). In further studies, we will analyze mental rotation performance in chronometric tests while controlling for hormone levels and self-concepts of primary school aged children to contribute to the discussion of psycho-social vs. biological–neuronal causes.
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
61
References Amorim, M. A., Isableu, B., & Jarraya, M. (2006). Embodied spatial transformations: “Body analogy” for the mental rotation of objects. Journal of Experimental Psychology, 135, 327–347. Archibald, A. B., Graber, J. A., & Brooks-Gunn, J. (2006). Pubertal processes and physiological growth in adolescence. In G. R. Adams, & M. D. Berzonsky (Eds.), Blackwell handbook of adolescence (2nd ed., pp. 24–47). Malden Blackwell Publishing. Casey, M. B., Colon, D., & Goris, Y. (1992). Family handedness as a predictor of mental rotation ability among minority girls in a math-science training program. Brain and Cognition, 18, 88–96. Courbois, Y. (2000). The role of stimulus axis in children’s ability to mentally rotate unfamiliar figures. European Journal of Cognitive Psychology, 12, 261–269. Heil, M., & Jansen, P. (2010). The relation between motor development and mental rotation ability in 5–6 years old children. European Journal of Developmental Science, 4, 66–74. Heller, K., & Geisler, H. (1983). KFT 1–3. Kognitiver Fähigkeits-Test (Grundschulform). Weinheim: Beltz Test GmbH. Huynh, H., & Feldt, L. S. (1976). Estimation of the Box correction for degrees of freedom from sample data in randomized block and split-plot designs. Journal of Educational Statistics, 1, 69–82. Jansen-Osmann, P., & Heil, M. (2007a). Suitable stimuli to obtain (no) gender differences in the speed of cognitive processes involved in mental rotation. Brain and Cognition, 64, 217–227. Jansen-Osmann, P., & Heil, M. (2007b). Developmental aspects of the laterality of ERP effects during mental rotation. Neuroreport, 18, 175–178. Jöckel, K.-H., Babitsch, B., Bellach, B.-M., Bloomfield, K., Hoffmeyer-Zlotnik, Winkler, J., et al. (1998). Messung und Quantifizierung soziodemografischer Merkmale in epidemiologischen Studien. In W. Ahrens, B.-M. Bellach, & K.-H. Jöckel (Eds.), Messung soziodemografischer Merkmale in der Epidemiologie. (pp. 7–38). München: MMV Münchner Medizin Verlag. Just, M. A., & Carpenter, P. A. (1985). Cognitive coordinate systems: Accounts of mental rotation and individual differences in spatial ability. Psychological Review, 92, 137–171. Kail, R. (1988). Developmental functions for speed of cognitive processes. Journal of Experimental Child Psychology, 45, 339–364. Kail, R. (1993). Processing time decreases globally at an exponential rate during childhood and adolescence. Journal of Experimental Child Psychology, 56, 254–265. Kail, R., Pellegrino, J., & Carter, P. (1980). Developmental changes in mental rotation. Journal of Experimental Child Psychology, 29, 102–116. Kosslyn, S. M., Margolis, J. A., Barrett, A. M., & Goldknopf, E. J. (1990). Age differences in imagery abilities. Child Development, 61, 995–1010. Krüger, M., & Krist, H. (2009). Imagery and motor processes – when are they connected? The mental rotation of body parts in development. Journal of Cognition and Development, 10, 239–261. Levine, S. C., Huttenlocher, J., Taylor, A., & Langrock, A. (1999). Early sex differences in spatial skill. Developmental Psychology, 35, 940–949. Levine, S. C., Ratliff, K. R., Huttenlocher, J., & Cannon, J. (2012). Early puzzle play. A predictor of preschoolers’ spatial transformation skill. Developmental Psychology, 48, 530–542. Levine, S. C., Vasilyeva, M., Lourenco, S. F., Newcombe, N. S., & Huttenlocher, J. (2005). Socioeconomic status modifies the sex difference in spatial skill. Psychological Science, 16, 841–845. Linn, M. C., & Peterson, A. C. (1985). Emergence and characterization of sex-differences in spatial ability: A meta-analysis. Child Development, 56, 1479–1498. Marmor, G. S. (1975). Development of kinetic images: When does the child first represent movement in mental images? Cognitive Psychology, 7, 548–559. Moè, A., & Pazzaglia, F. (2006). Following the instructions! Effects of gender beliefs in mental rotation. Learning and Individual Differences, 16, 369–377. Moore, D. S., & Johnson, S. P. (2008). Mental rotation in human infants: A sex difference. Psychological Science, 19, 1063–1066. Neuburger, S., Jansen, P., Heil, M., & Quaiser-Pohl, C. (2011). Gender differences in pre-adolescents’ mental rotation performance: Do they depend on grade and stimulus type? Personality and Individual Differences, 50, 1238–1242. Newcombe, N. (2002). The nativist-empiricist controversy in the context of recent research on spatial and quantitative development. Psychological Science, 13, 395–401. Newcombe, N., Bandura, M. M., & Taylor, D. G. (1983). Sex differences in spatial ability and spatial activities. Sex Roles, 9, 377–386. Newcombe, N., & Frick, A. (2010). Early education for spatial intelligence: Why, what and how. Mind, Brain, and Education, 4, 102–111. Perrucci, V., Agnoli, F., & Albiero, P. (2008). Children’s performance in mental rotation tasks: Orientation-free features flatten the slope. Developmental Science, 11, 732–742. Peters, M., & Battista, C. (2008). Applications of mental rotation figures of the Shepard and Metzler Type and description of a mental rotation stimulus library. Brain and Cognition, 66, 260–264. Peters, M., Laeng, B., Latham, K., Jackson, M., Zaiyouna, R., & Richardson, C. (1995). A redrawn Vandenberg and Kuse Mental Rotation Test: Different versions and factors that affect performance. Brain and Cognition, 28, 39–58. Quaiser-Pohl, C. (2003). The mental-cutting test “Schnitte” and the picture-rotation-test – two new measures to assess spatial ability. International Journal of Testing, 3, 219–231. Quaiser-Pohl, C., Rohe, A., & Amberger, T. (2010). The Solution strategy as an indicator of the developmental stage of pre-school children’s mental-rotation ability. Journal of Individual Differences, 31, 95–100. Quinn, P. C., & Liben, L. S. (2008). A sex differences in mental rotation in young infants. Psychological Science, 19, 1067–1070. Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects. Science, 171, 701–703. Snodgrass, J. G., & Corwin, J. (1988). Pragmatics of measuring recognition memory: Application to dementia and amnesia. Journal of Experimental Psychology: General, 117, 34–50. Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174–215.
62
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
Steele, C. (1997). A threat in the air – how stereotypes shape intellectual identity and performance. American Psychologist, 52, 613–629. Thurstone, T. G. (1958). Manual for the SRA primary mental abilities. Chicago: Science Research Associates. Titze, C., Jansen, P., & Heil, M. (2010). Mental rotation performance and the effect of gender in 4th graders and adults. European Journal of Developmental Psychology, 7, 432–444. Vandenberg, S. G., & Kuse, A. R. (1978). Mental rotations. A group test of three-dimensional spatial visualization. Perceptual and Motor Skills, 47, 599–604. Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in spatial abilities: A meta-analysis and consideration of critical variables. Psychological Bulletin, 117, 250–270. Wiedenbauer, G., & Jansen-Osmann, P. (2008). Manual training of mental rotation in children. Learning and Instruction, 18, 30–41.
Contents lists available at SciVerse ScienceDirect
Cognitive Development
Mental rotation performance in primary school age children: Are there gender differences in chronometric tests? P. Jansen a,∗, A. Schmelter a, C. Quaiser-Pohl b, S. Neuburger b, M. Heil c a b c
Universität Regensburg, Institute of Sport Science, Regensburg, Germany Universität Koblenz, Institute of Psychology, Koblenz, Germany Heinrich-Heine Universität Düsseldorf, Institute of Experimental Psychology, Düsseldorf, Germany
a r t i c l e Keywords: Mental rotation Pre-adolescence Gender differences Stimulus type
i n f o
a b s t r a c t In contrast to the well documented male advantage in psychometric mental rotation tests, gender differences in chronometric experimental designs are still under dispute. Therefore, a systematic investigation of gender differences in mental rotation performance in primary-school children is presented in this paper. A chronometric mental rotation task was used to test 449 second and fourth graders. The children were tested in three separate groups each with different stimulus material (animal drawings, letters, or cube figures). The results show that chronometric mental rotation tasks with cube figures – even rotated in picture plane only – were too difficult for children in both age groups. Further analyses with animal drawings and letters as stimuli revealed an overall gender difference in response time (RT) favoring males, an increasing RT with increasing angular disparity for all children, and faster RTs for fourth graders compared to second graders. This is the first study which has shown consistent gender differences in chronometric mental rotation with primary school aged children regarding reaction time and accuracy while considering appropriate stimuli. © 2012 Elsevier Inc. All rights reserved.
∗ Corresponding author at: Institute of Sport Science, Universitätsstr. 31, 93053 Regensburg, Germany. Tel.: +49 941 943 2518; fax: +49 941 943 4490. E-mail address: [email protected] (P. Jansen). 0885-2014/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cogdev.2012.08.005
52
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
1. Introduction Mental rotation is the ability to rotate two- or three-dimensional stimuli in the mind (Shepard & Metzler, 1971). The main goal of this study is to investigate gender differences in the chronometric mental rotation performance of school age children when using different types of stimuli. This is a combination of topics which has not been systematically investigated until now. 1.1. Mental rotation performance measured with psychometric and chronometric tests in children Psychometric mental rotation tests, where participants have to solve time-limited mental rotation tasks on a paper sheet, can be used to test children as young as 5 years old (Quaiser-Pohl, Rohe, & Amberger, 2010) when using age appropriate stimuli (Quaiser-Pohl, 2003). With the most widely used paper–pencil mental rotation test, the MRT (Peters et al., 1995; Vandenberg & Kuse, 1978), psychometric mental rotation performance can be reliably measured in primary school-aged children (9–11 years) (Titze, Jansen, & Heil, 2010). In the MRT, each of the 24 items consists of one target on the left side and four comparison items on the right side. Two of the four comparisons are “correct” (identical to the left item) and two are “incorrect”. Participants must choose both of the correct stimuli within a time limit of three minutes for 12 items. Accuracy is measured. In addition to psychometric tests there are chronometric mental rotation tests. In these tests, participants sit in front of a computer screen and are shown pairs of mental rotation objects. They must decide as fast and as accurately as possible if the objects are the same or are mirror images of each other by pressing a key. Accuracy and reaction time are measured. It has been demonstrated that children are able to solve chronometric mental rotation tasks even before entering school. Marmor (1975) showed that 5-year-old children can mentally rotate two-dimensional figures in the picture plane. However, they were twice as slow as 8-year-old children. These results were confirmed in other studies (e.g. Kosslyn, Margolis, Barrett, & Goldknopf, 1990), but they are also still controversially discussed (Newcombe, 2002). Results from additional studies have revealed a high individual variability and some studies have failed to replicate Marmor’s result (compare Newcombe & Frick, 2010). Kail, Pellegrino, and Carter (1980) investigated the development of mental rotation performance after the age of eight. They found a linear mental-rotation function using alphanumeric and abstract symbols and showed that mental-rotation speed nearly doubled from 8 years old to adulthood. Concerning the influence of stimulus type, Courbois (2000) showed that the ability to rotate unfamiliar stimuli with no salient axis improves from five to 8 years of age. Furthermore, Perrucci, Agnoli, and Albiero (2008) demonstrated that children at the age of 6 years old are already able to compare two stimuli at different orientations dependent on orientation-free features, such as the color of some part of the stimuli. In this case they are able to respond without using mental rotation. 1.2. The importance of gender differences in children measured with psychometric and chronometric tests Gender differences in mental-rotation tasks are widely discussed and well investigated in psychometric, paper–pencil tests (Peters & Battista, 2008). Using psychometric tests with three different stimulus conditions (animal pictures, letters, and cube figures) it was shown that fourth graders, but not second graders, showed a small, significant, stimulus-independent gender difference favoring males (Neuburger, Jansen, Heil, & Quaiser-Pohl, 2011). This is in line with another study showing a large gender difference favoring males only in older (mean age: 10.3 years) but not in younger fourth graders (mean age: 9.3 years) (Titze et al., 2010). These gender differences are often discussed on a psycho-social or biological–neuronal basis. Psycho-social theories often argue that gender differences are based on the influence of attitudes or stereotypes (Moè & Pazzaglia, 2006; Steele, 1997) or on different experiences based on gender. For example, Levine, Ratliff, Huttenlocher, and Cannon 2012) showed that the quality, but not the frequency, of puzzle play was higher for boys than for girls (between 2 and 4 years old) and that this variation predicts performance on spatial transformation tasks. Biological theories stress the importance of brain organization and hormonal (organizational and activational) influences. Both psycho-social and biological approaches are empirically supported
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
53
and their factors seem to interact in a complex way (Casey, Colon, & Goris, 1992). Furthermore, especially with children, task attributes in the mental rotation task itself also play an important role in mental rotation performance (e.g. Courbois, 2000; Perrucci et al., 2008). Gender differences in chronometric mental-rotation tasks have been investigated more thoroughly in adulthood than in childhood. Seven (5 with children under 14 years old) of the 15 chronometric studies in the meta-analysis of Voyer, Voyer, and Bryden (1995) showed no gender differences at all, resulting in a small to medium overall effect size (d = .37). However, studies are difficult to compare because they vary with respect to stimulus types, sample sizes, and the analyzed variables of information processing. In a systematic study with a total of N = 360 adult participants, Jansen-Osmann and Heil (2007a) demonstrated that only polygons, but not characters, animal drawings, PMA symbols (2D geometric objects from the Primary Mental Ability Battery, Thurstone, 1958), or cube figures, produced a substantial and reliable gender difference in mental-rotation speed. This suggests that the finding of a male advantage in psychometric tests cannot automatically be generalized to chronometric mentalrotation tasks. Studies concerning gender differences with children using chronometric tests are rare. Krüger and Krist (2009) used pictures of body parts as stimuli material and found an unexpected gender difference: kindergartner’s boys made more errors than girls. In another study, Heil and Jansen (2010) investigated gender differences in both psychometric and chronometric mental rotation tests and in a standardized math test for 7–8-year-old children (47 girls, and 62 boys). Gender differences favoring boys were only found in accuracy measurements (psychometric mental rotation test, math test, and error rate of the chronometric test). 1.3. Goal of this study Regarding the use of chronometric mental rotation tests, a systematic investigation regarding gender difference, as it was presented for adults (Jansen-Osmann & Heil, 2007b), is missing for school age children. The main goal of this study is to investigate the gender differences present in school age children using a chronometric test and to assess how appropriate different stimuli are for this age group. According to the study of Heil and Jansen (2010) a gender difference in a chronometric mental rotation test using animal drawings is anticipated for accuracy measurements. The study presented here will add to their findings because we are additionally investigating the influence of different types of stimuli and different age groups on chronometric mental rotation performance. The study presented here will add to their findings because we are additionally investigating the influence of different types of stimuli and different age groups on chronometric mental rotation performance. Additionally, general intelligence and socioeconomic status were controlled, because socioeconomic status has been shown to modify gender differences in spatial tasks (Levine, Vasilyeva, Lourenco, Newcombe, & Huttenlocher, 2005). Because cube figures are more difficult to process than animal drawings or letters, indicated by longer reaction times (e.g. Jansen-Osmann & Heil, 2007a), we expected a larger gender difference for the cube figures. Furthermore, we expected a better mental rotation performance from fourth graders compared to second graders due to an increase in general processing speed (e.g. Kail, 1993) and rotation rate (Kail et al., 1980) with age. 2. Method 2.1. Participants The participants in this study consisted of 449 elementary-school children (228 second graders, 221 fourth graders) from schools in the area of Bonn, Western Germany. All parents gave their written, informed consent. The sample included children from families with low (13.4%), middle (14.3%), high (17.2%), and very high (48.7%) socio-economic status (SES). 6.4% did not complete the SES questionnaire. The socioeconomic status was measured by a questionnaire based on the measure provided by Jöckel et al. (1998). In this questionnaire, parents of the participants were asked to indicate their school leaving certificate and professional degree. Following Jöckel et al. (1998), rank-ordered values were assigned to the different combinations of graduation levels.
54
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
To ensure that we only used data from children that did attempt to solve the mental rotation test as instructed, we excluded the data of children whose error rates in at least one of the experimental conditions exceeded 90%. This resulted in the exclusion of 19 children, 12 boys (9 second graders) and 7 girls (6 second graders. From the remaining 430 children, 212 children were second graders (mean age: 7.94 years, SD = .5, range: 6.83–9.92 years), and 218 were fourth graders (mean age: 10.06, years, SD = .54, range: 8.25–12.17 years). Each child received little presents (value: .5D ) for participation. Additionally, participating classes were given 4D per child for the class treasury. Non-parametric statistics (Mann–Whitney U-test) showed that the SES did not differ between boys and girls (p = .53), second and fourth graders (p = .67), or the three stimulus condition groups (p = .67). Because of this, we could exclude that possible gender differences are modified by the socioeconomic status. Each child was tested in one of the three stimulus types (animal drawings, letters, or cube figures). The following numbers of children were analyzed in the different stimulus types: “animal drawings”: 60 second graders (32 girls and 28 boys) and 67 fourth graders (31 girls and 36 boys); “letters”: 85 second graders (48 girls and 37 boys) and 77 fourth graders (39 boys and 38 girls); “cube figures”: 67 second graders (34 girls and 33 boys) and 74 fourth graders (37 boys and 37 girls). 2.2. Material The three stimulus conditions were completely equivalent besides the stimuli used. The “animal drawings” condition consisted of colored drawings of 20 different animals (bear, camel, raccoon, cow, crocodile, dog, donkey, elephant, goat, grizzly bear, horse, leopard, lion, monkey, pig, rhino, sheep, tiger, turtle, and zebra, respectively, from Snodgrass and Vanderwart (1980)). The “letters” condition consisted of 15 alphanumeric letters (e, F, g (twice but different fonts), j, k (twice but different fonts), L, l, n (twice but different fonts), P, q (three times but different fonts) R, r, s, S, z). In condition 3, we used “cube figures” similar to those by Shepard and Metzler (1971). An example of each stimulus type is given in Fig. 1. In order to parallelize task demands for all three stimuli types and to avoid floor effects in secondgrade participants, we only used picture-plane rotations for the cube figures. Each stimulus had a maximal size of 7 by 7 cm on the screen with a space of 14 cm in between. The children were free to choose the most comfortable viewing distance. In each stimulus condition, the right stimulus was either identical to the left or mirror-reversed. The left stimulus was always presented upright. The angular disparity between the two stimuli was 45◦ , 90◦ , or 135◦ (clockwise or counter clockwise). With “animal drawings” and “letters” as stimuli each child had to complete 240 trials each. Because mental rotation with cube figures as the stimulus material was supposed to be much more difficult, each child only had to complete 120 trials to avoid overstraining the participants. After 20 trials a short break was administered in all stimuli conditions. The experiment was run on a laptop with a 15 monitor and lasted approximately 1 h. General cognitive abilities were controlled by administering the subtest “Reasoning” of the “Cognitive-Ability Test” (KFT 1–3, Heller & Geisler, 1983). This test consists of 15 items, which are made up of five pictures, respectively. Four of the five pictures belong to the same category (e.g. vegetable). The picture that does not fit into this category has to be detected and marked by the child. There is no time constraint. The test was chosen because it measures general non-verbal cognitive abilities, and because performance in this test does not depend on visual-spatial abilities. Furthermore, it can be administered in groups. 2.3. Procedure Children were tested in a quiet room during regular school time in groups of two to five children. The children were separated in this room so that every child was able to solve the tasks individually on a laptop. After a short introduction, children were given the KFT subtest. Since the KFT is a power test, there was no temporal constraint. When all children had finished with the KFT, the mental rotation task started. Each session was preceded by 40 unrecorded practice trials. Children were allowed to choose their own pace by pressing a key when they wanted to start the next run of 20 trials. Each trial started with the presentation of a 500 ms background gray screen. After this, the stimuli pair
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
55
Fig. 1. Sample items of the mental rotations tasks (animal drawings, letters, and cube figures).
was presented. Children had to decide as fast as possible if the stimuli were the “same” or “different” (mirror reversed). They were instructed to press the green marked mouse button when the stimuli were the “same” and the red marked mouse button when the stimuli were “different”. For pressing the key they were allowed to choose which finger to respond with. The children received feedback in the form of a “+” for correct responses and a “−” for incorrect responses only in the 40 training trials. Both feedbacks appeared for 500 ms in the center of the screen. The next trial began after 1500 ms. No feedback was given in the non-training trials. 3. Results For the statistical analysis of response time (RT), only trials with correct responses were used. RTs more than 2 SDs above or below the mean per condition and per participant were excluded, resulting in the exclusion of a total of 6.2% of the data. Analyses with untrimmed data revealed identical results. For all analyses clockwise and counter clockwise angular disparities were averaged. 3.1. Reasoning score Reasoning ability was defined as the number of correctly answered items in the KFT subtest and transformed into age-corrected standard values. An ANOVA with the KFT-score as a dependent variable
56
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
and the between subject factors “grade” (grade 2 vs. grade 4), “gender” (boys vs. girls), and “stimulus type” (animal drawings vs. letters vs. cube figures) was conducted. There were significant effects of the factors “gender”, F(1,418) = 13.60, p < .001, 2 = .03, and “grade”, F(1,418) = 35.02, p < .001, 2 = .08, but no significant main effect of the factor “stimulus type”, (F(2,418) = 2.05, n.s. The children in the 2nd grade (M = 47.34, SE = .64) had a lower score than the children in the 4th grade (M = 52.68, SE = .63), and boys (M = 48.34, SE = .64) had a lower score than girls (M = 51.67, SE = .63). The interaction between both factors, F(1,418) = 3.76, p = .053, 2 = .01, did not reach significance. In the 2nd grade, girls (M = 49.29, SE = .90) had a higher score than boys (M = 44.85, SE = .99), F(1,210) = 13.50, p < .001, 2 = .06. The score did not differ between girls (M = 53.59, SE = .87) and boys (M = 51.91, SE = .81) in the 4th grade, F(1,216) = 2.01, n.s. The KFT-score was considered as a covariate in the following analyses. 3.2. Accuracy score An error measurement was calculated for each participant. In reference to memory research (Snodgrass & Corwin, 1988), the PR-score (which is the abbreviation for the discrimination index according to Snodgrass and Corwin (1988)) was calculated for each angular disparity (45◦ , 90◦ , 135◦ ). The PR score is defined as the difference between the hits percentage (percentage of “same” responses for trials where “same” was the correct response) and the false alarm percentage (percentage of “same” responses for trials where “same” was the incorrect response). This measurement allows for the correction of guessing by chance and of guessing by always pressing the same button. An ANCOVA was conducted covarying the KFT score with the dependent measure “PR-score”, the within subject factor “angular disparity” (45◦ , 90◦ , 135◦ ), and the between subject factors “stimulus type” (animal drawings vs. letters vs. cube figures), “gender” (boys vs. girls), and “grade” (grade 2 vs. grade 4). The significance levels were corrected according to Huynh and Feldt (1976) in order to compensate for the non-sphericity of the data. The ANCOVA revealed a significant main effect for four factors: “angular disparity“, F(2, 834) = 5.28, p < .01, 2 = .01; “grade”, F(1,417) = 14.24, p < .001, 2 = .03; “gender”, F(1,417) = 11.52, p < .01, 2 = .03; and “stimulus type”, F(2,417) = 150.27, p < .001, 2 = .42. Furthermore, there was a significant interaction between the factors “grade” and “stimulus type”, F(2,417) = 7.37, p < .01, 2 = .03, between the factors “angular disparity” and “stimulus type”, F(4,834) = 3.60, p < .05, 2 = .02, and a three-way interaction between “angular disparity”, “stimulus type”, and “gender”, F(4,834) = 3.63, p < .05, 2 = .02 (see Fig. 2a and b).1 In Fig. 2a the accuracy score for the stimulus type “animal drawings” is presented showing a significant main effect of the factor “gender”, F(1,124) = 5.83, p < .05, 2 = .05, but not for “angular disparity”, F(1,124) = 3.00, n.s. However it does show a significant interaction between both factors, F(2,248) = 5.46, p < .05, 2 = .04. There was neither a gender difference for an angular disparity of 45◦ , F(1,125) = .13, n.s., nor for an angular disparity of 90◦ , F(1,125) = 2.58, n.s., but there was a gender difference for an angular disparity of 135◦ , F(1,125) = 5.32, p < .05, 2 = .04. Fig. 2b shows the accuracy score for the stimulus type “letters”. There was a significant main effect for “gender”, F(1,159) = 5.18, p < .05, 2 = .03 and “angular disparity”, F(2,318) = 8.16, p < .001, 2 = .05, 1 Because the PR-score is rarely used in the analysis of the accuracy rate in mental rotation the results were compared to the often used measurements of accuracy rate in mental rotation: analyzing only the accuracy rate for the responses on items which were the same showed a result which was comparable with the PR-score measurement regarding the main effects. The ANCOVA for the accuracy rate on same items revealed a significant main effect for four factors: “angular disparity”, F(2,834) = 13.82, p < .001, 2 = .03; “grade”, F(1,417) = 9.65, p < .01, 2 = .02; “gender” F(1,417) = 5.78, p < .05, 2 = .02; and “stimulus type”, F(2,417) = 33.96, p < .001, 2 = .14. Furthermore, there was a significant interaction between the factors “grade” and “stimulus type”, F(2,417) = 4.43, p < .01, 2 = .02, and a three-way interaction between “angular disparity”, “stimulus type”, and “grade”, F(4,834) = 2.55, p < .05, 2 = .01. When analyzing the accuracy rate for the responses on items which were different, the ANCOVA revealed a result which was comparable of the ANCOVA of the PR-score. A main effect of angular disparity was missing despite the significant interaction between angular disparity and KFT-score, F(2, 834) = 3.23, p < .05, 2 = .008. Furthermore, there were significant main effects of “grade”, F(1,417) = 9.18, p < .01, 2 = .02; “gender” F(1,417) = 9.39, p < .05, 2 = .02; and “stimulus type”, F(2,417) = 182.89, p < .001, 2 = .47. Additionally there were interactions between “grade” and “stimulus type”, F(2,417) = 5.39, p < .01, 2 = .03 and between “angular disparity”, “stimulus type”, and “gender”, F(4,834) = 2.91, p < .05, 2 = .01.
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
57
Fig. 2. PR-score (means and standard deviation) dependent on stimulus type, angular disparity, and gender. Part (a) shows the PR-score for animal drawings; part (b) shows the PR-scores for letters.
but no interaction between both factors, F(2,318) = 1.3, n.s. Boys (74.53%, SE = 2.93) had a greater accuracy score than girls (70.35%, SE = 3.00). The accuracy score was higher for an angular disparity of 45 ◦ (M = 80.07%, SE = 2.03), compared to 90 ◦ (M = 73.22%, SE = 2.24), F(1,161) = 46.46, p < .001, 2 = .23, compared to 135◦ (M = 63.64%, SE = 2.33), F(1,161) = 77.61, p < .001, 2 = .33. Concerning the accuracy score for the stimulus type “cube figures” there was no significant main effect of “angular disparity”, F(2,276) = .34, n.s., or “gender”, F(1,138) = 2.03, n.s., nor a significant interaction between both factors, F(2,276) = .76, n.s. The descriptive effect of angular disparity did not reach significance because the KFT-score was used as a covariate, F(2,276) = 3.84, p < .05, 2 = .03. The
58
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
Fig. 3. Reaction time (means and standard deviation) dependent on stimulus type and grade.
performance in the cube condition (averaged over all angular disparities) was beneath chance for girls (M = 37.78%, SE = 3.46) as well as for boys (M = 35.56%, SE = 3.43). 3.3. Response time Response time data of two children were excluded because their reaction time was more than three standard deviations above the mean. Because of the very poor PR-score for the stimulus type “cube figures” (see Section 3.2) further analyses concerning response time were only conducted with the stimuli “animal drawings” and “letters”. For these two stimulus types, all children who had an error rate lower than 30% in one of the six conditions (3 levels of angular disparity: 45◦ , 90◦ , 135◦ ; 2 types of stimuli presentation: identical or mirror-reversed) were included in further analysis (43 girls and 43 boys in grade 2, and 53 girls and 64 boys in grade 4). The cut-off criterion was decided based on the work of Perrucci et al. (2008), who only allowed children to participate in a mental rotation task when they had solved a pre-test stimulus matching task with less than 18% errors, and on the work of Wiedenbauer and Jansen-Osmann (2008) who excluded all children with an error rate over 40%. The KFT reasoning score was used as a covariate in the analysis of the response time. This analysis of covariance with response time as the dependent variable was restricted to responses to identical items only, because angular disparity is not unequivocally defined for responses to mirror-reversed items. The within subject factor was defined as “angular disparity” (45◦ , 90◦ , 135◦ ), and the between subject factors were defined as “gender” (boys vs. girls), “grade” (grade 2 vs. grade 4), and “stimulus type” (letters vs. animal drawings). The significance levels were corrected according to Huynh and Feldt (1976) in order to compensate for the non-sphericity of the data. Reliability of the data was given by measuring the split-half reliability using the odd-even method and the Spearman–Brown prediction formula for each grade and each angular disparity (grade 2: 45◦ : r = .72, 90◦ : r = .8, 135◦ : r = .83; grade 4: 45◦ : r = .77, 90◦ : r = .51, 135◦ : r = .86). The ANCOVA showed significant main effects of the factors “gender”, F(1,193) = 10.36, p < .005, 2 = .05, “grade”, F(1,193) = 32.98, p < .001, 2 = .15, and “angular disparity”, F(2,386) = 39.65, p < .001, 2 = .17, but not for “stimulus type”, F(1,193) = .81, n.s. (see Fig. 3). There was no significant interaction. Girls (M = 2000.30 ms, SE = 52.94) needed more time to solve the task than boys (M = 1817.68, SE = 45.98). Children in the 2nd grade (M = 2147.26 ms, SE = 55.19) showed longer response times than
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
59
children in 4th grade (M = 1720.79 ms, SE = 38.80), and furthermore, the response time increased with increasing angular disparity (45◦ : M = 1505.68 ms, SE = 30.99; 90◦ : M = 1839.72 ms, SE = 36.14; 135◦ : M = 2368.01 ms, SE = 47.00). 4. Discussion This is the first study which investigates the influence of grade, gender, and stimulus type, and their possible interactions, on the chronometric mental-rotation performance of elementary-school children. Previously these elements have only been explored in psychometric tests. Our first hypothesis regarding gender differences could only be partially confirmed: with animal drawings boys had a higher accuracy rate only at 135◦ of rotation; with letters the overall better accuracy rate of boys could be shown at all angles. Secondly, it was shown that boys were faster than girls at all rotational angles, and fourth graders were faster than second graders. Furthermore, the results show that most children in the 2nd grade and the 4th grade had substantial difficulties to solve a chronometric mental rotation test with cube figures rotated in the picture plane. Because of this, our hypothesis that cube figures lead to a larger gender difference could not be investigated in detail. 4.1. Gender differences in accuracy score Concerning accuracy rate, the results are different for letters and animals drawings. With animal drawings boys had a higher accuracy rate only at 135◦ of rotation; with letters an overall better accuracy rate of boys was demonstrated. The gender difference in accuracy score with animal drawings only at the higher angular disparity of 135◦ could be explained if the boys used a different strategy than the girls to solve these trials. Error rates are more sensitive than reaction time to the effect of holistic vs. piecemeal rotation. According to different sources of error in the matching process accuracy degrades as a function of angular disparity when adopting a piecemeal strategy, where participants compare each part of the figure with the corresponding parts of the reference object (Amorim, Isableu, & Jarraya, 2006). The higher error rate for girls with animal drawings at 135◦ suggests that more girls may use a piecemeal strategy while rotating animal drawings. Another explanation might be that the animal drawings are “embodied”. This strategy involves mapping the body axes (head–feet, front–back, and left–right) onto the animal drawing. If this strategy was used it might be assumed that girls had more difficulties to rotate the picture into the unfamiliar body positions, such as 135◦ compared to 45◦ and 90◦ . This could be due to the girls having less motor experience with these angles than the boys, but this remains rather speculative and has to be investigated further. Boys showed an advantage in accuracy score for letters compared to girls at all angular disparities. This might indicate that the strategies between boys and girls did not differ at either age while solving a mental rotation task with letters. Differentiating which strategy is used between boys and girls could be investigated with eye-tracking analysis (Just & Carpenter, 1985). 4.2. Gender differences in reaction time Concerning the reaction time when using letters and animal drawings as stimulus material, we found an increasing reaction time with increasing angular disparity, a longer reaction time for 2nd graders than for 4th graders, and a longer reaction time for girls than for boys. The decreasing reaction time from grade 2 to grade 4 is in accordance with a general increase in the speed of cognitive processes with age (Kail, 1988). The increasing reaction time with increasing angular disparity is in line with many other studies including the classic Shepard and Metzler (1971) study. Interestingly, this increase, which is assumed to reflect the process of mentally rotating the image, did not differ between boys and girls. Therefore, the effect of gender might be due to perceptual or motor processes executed before and after the rotation itself. In further studies a 0◦ condition has to be included to differentiate between the mental rotation process and the perceptual and motor processes in the mental rotation task. The obtained gender differences in this study are in accordance with Linn and Peterson (1985) who assumed that there is an emergence of gender difference favoring males as early as children can
60
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
mentally rotate. However, the meta-analysis of Linn and Peterson (1985) only included studies in which the psychometric MRT was used. Therefore, the revealed gender differences in chronometric mental-rotation tasks in primary school aged children are a new result. With regard to the influence of gender, these results are in contrast to the results of chronometric studies which used the same chronometric stimulus material with adults (Jansen-Osmann & Heil, 2007a). In their study no gender differences could be detected with letters and animal drawings as stimuli types. One possible explanation is that the different designs might contribute to these conflicting results: In the study with adults, trials were presented in blocks “always upright” or “sometimes upright” with angular disparities of 0◦ , 90◦ , and 180◦ , whereas in this study they were presented randomized and with angular disparities of 45◦ , 90◦ , and 135◦ . Another difference is that the participants in the adult study received feedback after each trial whereas the children only received feedback during the practice trials. While the women were able to see their “good” performance the girls did not.
4.3. Cube figures as stimuli A new result of the present study is that children were not able to solve the chronometric mentalrotation task with cube figures. To our knowledge there is no study investigating mental rotation in primary school-age children with cube figures as stimuli. For example, Kail et al. (1980) used alphanumeric characters and abstract symbols, Marmor (1975) used drawings rotated in picture plane, and Jansen-Osmann and Heil (2007b) used letters to investigate the neuronal correlates of the mental rotation process in primary school-aged children. Results are in accordance with the study of JansenOsmann and Heil (2007a) who showed that the mental rotation speed in adults was slower for cube figures than for letters, animal drawings, PMA-figures, and polygons. This means that adults also found it more difficult to solve a mental rotation task with cube figures than with other stimulus material. Children are less familiar with cube figures than with letters and animal drawings, this could cause the cube figures to be more difficult to process. These results suggest that the mental rotation performance in general is stimuli dependent and the stimuli used must be considered as a relevant factor in developmental and differential visual spatial research.
5. Conclusions On the basis of the present results, we can conclude that gender differences in chronometric mentalrotation tests favoring males deserve attention when letters and animal drawings are used as stimuli and that cube figures are too difficult for children between the ages of 8–10 years to process. This is a new finding because in previous chronometric studies gender differences are random (e.g. Voyer et al., 1995) or do only appear in accuracy measurements (Heil & Jansen, 2010). Finally, we know that our data will contribute to the discussion of gender differences in chronometric mental-rotation tests and reveal that experimental designs must be carefully considered when comparing studies of gender differences in mental rotation performance at this age. This study focused on children in middle childhood, but it is important to say that gender differences in spatial transformation task are found much earlier in life as demonstrated in preschool children (Levine, Huttenlocher, Taylor, & Langrock, 1999) and infants (Moore & Johnson, 2008; Quinn & Liben, 2008). According to our results the obtained gender differences in children younger than 8 years old might be also investigated regarding the different designs used. With this study we could not differentiate between a psycho-social and a biological–neuronal hypothesis. One concern is that “hormone levels begin to rise in middle childhood” (Archibald, Graber, & Brooks-Gunn, 2006, p. 25), and another concern is that, stereotypic conceptions of gender develop in middle childhood (Newcombe, Bandura, & Taylor, 1983). In further studies, we will analyze mental rotation performance in chronometric tests while controlling for hormone levels and self-concepts of primary school aged children to contribute to the discussion of psycho-social vs. biological–neuronal causes.
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
61
References Amorim, M. A., Isableu, B., & Jarraya, M. (2006). Embodied spatial transformations: “Body analogy” for the mental rotation of objects. Journal of Experimental Psychology, 135, 327–347. Archibald, A. B., Graber, J. A., & Brooks-Gunn, J. (2006). Pubertal processes and physiological growth in adolescence. In G. R. Adams, & M. D. Berzonsky (Eds.), Blackwell handbook of adolescence (2nd ed., pp. 24–47). Malden Blackwell Publishing. Casey, M. B., Colon, D., & Goris, Y. (1992). Family handedness as a predictor of mental rotation ability among minority girls in a math-science training program. Brain and Cognition, 18, 88–96. Courbois, Y. (2000). The role of stimulus axis in children’s ability to mentally rotate unfamiliar figures. European Journal of Cognitive Psychology, 12, 261–269. Heil, M., & Jansen, P. (2010). The relation between motor development and mental rotation ability in 5–6 years old children. European Journal of Developmental Science, 4, 66–74. Heller, K., & Geisler, H. (1983). KFT 1–3. Kognitiver Fähigkeits-Test (Grundschulform). Weinheim: Beltz Test GmbH. Huynh, H., & Feldt, L. S. (1976). Estimation of the Box correction for degrees of freedom from sample data in randomized block and split-plot designs. Journal of Educational Statistics, 1, 69–82. Jansen-Osmann, P., & Heil, M. (2007a). Suitable stimuli to obtain (no) gender differences in the speed of cognitive processes involved in mental rotation. Brain and Cognition, 64, 217–227. Jansen-Osmann, P., & Heil, M. (2007b). Developmental aspects of the laterality of ERP effects during mental rotation. Neuroreport, 18, 175–178. Jöckel, K.-H., Babitsch, B., Bellach, B.-M., Bloomfield, K., Hoffmeyer-Zlotnik, Winkler, J., et al. (1998). Messung und Quantifizierung soziodemografischer Merkmale in epidemiologischen Studien. In W. Ahrens, B.-M. Bellach, & K.-H. Jöckel (Eds.), Messung soziodemografischer Merkmale in der Epidemiologie. (pp. 7–38). München: MMV Münchner Medizin Verlag. Just, M. A., & Carpenter, P. A. (1985). Cognitive coordinate systems: Accounts of mental rotation and individual differences in spatial ability. Psychological Review, 92, 137–171. Kail, R. (1988). Developmental functions for speed of cognitive processes. Journal of Experimental Child Psychology, 45, 339–364. Kail, R. (1993). Processing time decreases globally at an exponential rate during childhood and adolescence. Journal of Experimental Child Psychology, 56, 254–265. Kail, R., Pellegrino, J., & Carter, P. (1980). Developmental changes in mental rotation. Journal of Experimental Child Psychology, 29, 102–116. Kosslyn, S. M., Margolis, J. A., Barrett, A. M., & Goldknopf, E. J. (1990). Age differences in imagery abilities. Child Development, 61, 995–1010. Krüger, M., & Krist, H. (2009). Imagery and motor processes – when are they connected? The mental rotation of body parts in development. Journal of Cognition and Development, 10, 239–261. Levine, S. C., Huttenlocher, J., Taylor, A., & Langrock, A. (1999). Early sex differences in spatial skill. Developmental Psychology, 35, 940–949. Levine, S. C., Ratliff, K. R., Huttenlocher, J., & Cannon, J. (2012). Early puzzle play. A predictor of preschoolers’ spatial transformation skill. Developmental Psychology, 48, 530–542. Levine, S. C., Vasilyeva, M., Lourenco, S. F., Newcombe, N. S., & Huttenlocher, J. (2005). Socioeconomic status modifies the sex difference in spatial skill. Psychological Science, 16, 841–845. Linn, M. C., & Peterson, A. C. (1985). Emergence and characterization of sex-differences in spatial ability: A meta-analysis. Child Development, 56, 1479–1498. Marmor, G. S. (1975). Development of kinetic images: When does the child first represent movement in mental images? Cognitive Psychology, 7, 548–559. Moè, A., & Pazzaglia, F. (2006). Following the instructions! Effects of gender beliefs in mental rotation. Learning and Individual Differences, 16, 369–377. Moore, D. S., & Johnson, S. P. (2008). Mental rotation in human infants: A sex difference. Psychological Science, 19, 1063–1066. Neuburger, S., Jansen, P., Heil, M., & Quaiser-Pohl, C. (2011). Gender differences in pre-adolescents’ mental rotation performance: Do they depend on grade and stimulus type? Personality and Individual Differences, 50, 1238–1242. Newcombe, N. (2002). The nativist-empiricist controversy in the context of recent research on spatial and quantitative development. Psychological Science, 13, 395–401. Newcombe, N., Bandura, M. M., & Taylor, D. G. (1983). Sex differences in spatial ability and spatial activities. Sex Roles, 9, 377–386. Newcombe, N., & Frick, A. (2010). Early education for spatial intelligence: Why, what and how. Mind, Brain, and Education, 4, 102–111. Perrucci, V., Agnoli, F., & Albiero, P. (2008). Children’s performance in mental rotation tasks: Orientation-free features flatten the slope. Developmental Science, 11, 732–742. Peters, M., & Battista, C. (2008). Applications of mental rotation figures of the Shepard and Metzler Type and description of a mental rotation stimulus library. Brain and Cognition, 66, 260–264. Peters, M., Laeng, B., Latham, K., Jackson, M., Zaiyouna, R., & Richardson, C. (1995). A redrawn Vandenberg and Kuse Mental Rotation Test: Different versions and factors that affect performance. Brain and Cognition, 28, 39–58. Quaiser-Pohl, C. (2003). The mental-cutting test “Schnitte” and the picture-rotation-test – two new measures to assess spatial ability. International Journal of Testing, 3, 219–231. Quaiser-Pohl, C., Rohe, A., & Amberger, T. (2010). The Solution strategy as an indicator of the developmental stage of pre-school children’s mental-rotation ability. Journal of Individual Differences, 31, 95–100. Quinn, P. C., & Liben, L. S. (2008). A sex differences in mental rotation in young infants. Psychological Science, 19, 1067–1070. Shepard, R. N., & Metzler, J. (1971). Mental rotation of three-dimensional objects. Science, 171, 701–703. Snodgrass, J. G., & Corwin, J. (1988). Pragmatics of measuring recognition memory: Application to dementia and amnesia. Journal of Experimental Psychology: General, 117, 34–50. Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174–215.
62
P. Jansen et al. / Cognitive Development 28 (2013) 51–62
Steele, C. (1997). A threat in the air – how stereotypes shape intellectual identity and performance. American Psychologist, 52, 613–629. Thurstone, T. G. (1958). Manual for the SRA primary mental abilities. Chicago: Science Research Associates. Titze, C., Jansen, P., & Heil, M. (2010). Mental rotation performance and the effect of gender in 4th graders and adults. European Journal of Developmental Psychology, 7, 432–444. Vandenberg, S. G., & Kuse, A. R. (1978). Mental rotations. A group test of three-dimensional spatial visualization. Perceptual and Motor Skills, 47, 599–604. Voyer, D., Voyer, S., & Bryden, M. P. (1995). Magnitude of sex differences in spatial abilities: A meta-analysis and consideration of critical variables. Psychological Bulletin, 117, 250–270. Wiedenbauer, G., & Jansen-Osmann, P. (2008). Manual training of mental rotation in children. Learning and Instruction, 18, 30–41.