Adult Gender and Hand Preference Differences in Eye-Hand Visual Reaction Times

Left hand writing the German word "Linksh...
Image via Wikipedia


This research study investigated whether eye-hand reaction times (RT) can be related to differences in gender and handedness. The assumption that prompted this study sought to address the (controversial) question of whether masculinity and/or left-handedness might have an effect on adult reaction times over and above those usually found due to individual differences. In the study, reaction time data were collected from forty adult participants (N=40, M/F ratio = 1:1, RH ratio = 1:1) between 22-50yrs using a standardised format reaction time test measurement software application. Participants’ mean simple reaction time scores (msecs) were calculated based upon a sample of one hundred individual simple reaction time tests per participant (i.e. total sample of 4000 reaction times).

Participants were given thirty practice reaction time tests. Practice was followed by one hundred simple reaction time tests. From the total trial sample of simple reaction time scores for each participant mean simple reaction times were found. Mean simple reaction times across gender and handedness were then analysed by comparison of means and between-subject two-way analysis of variance (2×2 ANOVA). In addition, a separate one-way analysis of variance looked at whether age (years) might also be an affective factor in the study.

The study did not find evidence to support the suggestion that adult males performed significantly better than adult females in mean simple reaction times. Findings did appear consistent, however, with other larger studies which have suggested that left-handers results demonstrate a significant and robust positive main effect in reaction time when compared to right-handers.


How do gender and handedness differences relate to measurable events in eye-hand visual reaction time responsiveness? Moreover, to what extent are such differences, if reliable, indicative of real gender and handedness relatedness? In basic research into reaction times, one is probably conducting investigations into cognitive phenomena more or less as old as experimental psychology itself (e.g. Donders, 1868; Freud, 1895; Galton, 1899; Helmholtz, 1867; Jastrow, 1890; Külpe, 1895, Merkel, 1885, Wundt, 1880). Thus, the pursuit of the investigation of response timing has been an indicative meter of human cognitive processing through the utility of reaction time (chronometry). Voluntary and involuntary response measures alike, can be said to have gained purchase and burgeoned forth from Donders’ (1868) foundational studies (e.g. Cooper & Shepard, 1973; Fitt, 1954; Libet, 1965, 1979; Hicks, 1952; Posner, 1978, 2005; Shepard & Metzler, 1971; Sternberg, 1969; see also Ratcliff & Smith, 2004; Sanders, 1998; Welford, 1980; Whelan, 2008).

Notwithstanding, the substantial corpus of work that exists under the various headings of cerebral dominance or handedness or lateral dominance or hemispheric specialisation (>7630 articles between 1999-2009 found using Scholar), relatively few research studies over the last decade have related gender and handedness differences in the basic research of eye-hand visual reaction time responsiveness. This actuality is not for lack of research in the field, but owes more perhaps to a migration toward a methodological preference for meta-analyses largely due to the unwieldy volume of evidence available (e.g. Duval & Tweedie, 2000; Egger & Smith, 1997; Glass, 1976; Higgins, Thompson, Deeks, & Altman, 2003; Horrey & Wickens, 2006; Hunter & Schmidt, 1990; Lalumiere, Blanchard, & Zucker, 2000; Oyserman, Coon, & Kemmelmeier, 2002; Papadatou-Pastou, Martin, Munafò, & Jones, 2008; Rosenthal & DiMatteo, 2001; Silverman, 2006; Thompson & Higgins, 2002; Verhaeghen & Salthouse, 1997).

Simple Reaction Time

Basic research investigating differences in eye-hand visual reaction time in adults’ gender and handedness use the difference between the onset of stimuli and the associated voluntary response as a measure of reaction, customarily shown in milliseconds. In a psychological literature review of reaction timing, overarching suggestions have been made that appear to point toward a contention (a) males are more prone to left-handedness (see below), and, (b) that handedness can effect cognitive ability (see below). It is clear that potential questions, however, obtaining from these proposed inequalities in reaction time, and, therein, a surmised inequality in mentation abilities between the genders, might easily be mistakenly taken as kindling for the type of suppositions never too far from those encountered in the fundamental attribution errors of common experience and folk psychology alike. Thus, at the outset of this study the research sought to be inclusive toward the ramifications of whether studies concerned with chronometry by gender might be reasonably adjudged as having a quantitative evidential basis for controlling for handedness (e.g. Crow, Crow, Done, & Leask, 1998). It may be noteworthy that the telos for any demarcation of gender difference may also be reasonably expected to have a much wider cultural bearing not simply restricted to the experimental psychological community.

That having been said, voluntary reaction times (>120msec) are known to be significantly slower than are reflex responses (>40msec) elicited by comparable stimuli (see also Kandel, Schwartz, & Jessell, 2000). It is also widely accepted that mean reaction time to auditory stimulus (>120-140msec) are faster than the mean reaction time to visual stimulus (>150-180msec) (e.g. Brebner & Welford, 1980; Kandel, Schwartz, & Jessell, 2000; Welford, 1980). Indeed, this differential between audition and vision might be attributed to a multifactorial combination of synaptic complexity and neural conduction distance, as the mean reported response to auditory stimulus in adults takes between 8-10msec (Kemp, 1973) to reach the brain, whereas mean reported response to visual stimulus in adults takes between 20-40msec (Marshall, Talbot, & Ades, 1943). In separate findings put forward by Kohfield (1971) and Luce (1986) whilst investigating stimulus intensity, each have shown that sufficiently high or low stimulus intensity (i.e. luminosity, loudness) might serve to negate the differential between response times to auditory and visual stimulus. It is perhaps noteworthy that the differential between auditory and visual reaction times appears to persist whether a participant is requested to make a simple reaction time or choice reaction time response (Sander, 1998). Since the earliest basic psycho-physiological experiments in the 1890s, researchers have been able to show repeatedly that responsiveness increases non-linearly with the increasing availability of choice (e.g. Laming, 1968; Luce, 1986; Sternberg, 1969). Nevertheless, more contemporary thought holds to a more subtle view that, whilst complexity does indeed increase reaction time, as the participant learns to anticipate a correct response, so too a non-linear decrease in reaction time will follow (the choice effect) without a direct relation to the complexity of the original task (Ando, Kida, & Oda, 2002, 2004; Nissen & Bullmer, 1987). A hypothesis to be tested in future neuropsychological research might be reasonably imagined to take place in laboratories capable of looking more closely than we can presently into the choice effect in relation to the natural capacity for coextensive serial and parallel cognitive processing in the brain, perhaps even at the dendritic level.

More specifically of interest to this study, Hsieh, Lin, & Chen (2007) recently conducted experiments utilising a simulated vibration of the display during computerised visual reaction time tests. Hsieh, Lin, & Chen (2007) found that not only did reaction time increase with perceived vibration increase, but also error rates per sample increased as well as a greater incidence of reported fatigue among participants. It would appear, then, that the suggestions made by Hsieh, Lin, & Chen (2007) for the effects of display vibration might be of utility to the majority of researchers investigating reaction time responsiveness, not least those conducting out-of laboratory experiments.

As mentioned previously, the present study into gender and hand preference differences in mean simple reaction times used an equally balanced (N=40, M/F ratio = 1:1, RH/LH ratio = 1:1) sample drawn from the adult population (22-50yrs, Mean = 31.3yrs, STD = 9.7) to measure the extent to which gender and handedness effectively produce differences in mean simple reaction times. The study sought to investigate the relevance of the hypothesis held by some researchers that masculinity (e.g. Adam, Paas, Buekers, Wuyts, Spijkers & Wallmeyer, 1999; Dane & Erzurumluoğlu, 2003; Der & Deary, 2006; Noble, Baker & Jones, 1964; Welford, 1980) and/or left-handedness (e.g. Bartélémy & Boulinquez, 2002, 2001; Boulinquez & Bartélémy, 2000; Bryden, 2002; Dane & Erzurumluoğlu, 2003; Miller & Van Nes, 2007; Peters & Ivanoff, 1999) might have a positive and strong effect on reduced simple reaction times.

The research strategy aimed to follow protocols sourced from other reaction time studies (e.g. Fieandt, Huhtala, Kullberg, & Saarl, 1956; Luce, 1986; Sanders, 1998; Welford, 1980; Whelan, 2008), in that, after a short practice period, average simple reaction time scores were taken from the result of the sum of the population of reaction trials per individual (i.e. 100 trials) divided by the number of reaction trials (i.e. ∑ of population mean reaction time scores / 100 trials = Mean reaction time (msecs)). Each participants’ (N=40) mean simple reaction time were then analysed by gender and handedness (2×2 ANOVA).

Gender & Handedness (Hand preference)

Contemporary cognitive psychological thought holds that handedness is something that we are born with rather than something acquired by experience. Ensuing debate appears to have constellated around whether right-handedness is either a population bias (e.g. Annett, 2002; McManus, 1991), a genetic predisposition (e.g. Annett, 1985; Corballis, 1997; Jones & Martin, 2000; Klar, 1996; McManus, 1985; McManus & Bryden, 1992), a developmental relation to intrauterine factors (e.g. Hepper, McCartney, & Alyson, 1998; Previc, 1991), or even an anthropological concern with its roots in a distant migration of hand preference (e.g. Probiner, 1999; Toth, 1985; Uhrbrock, 1973). Indeed, gender differences in handedness researchers have also reported on the high differential incidence rates of mixed- and left-handedness reaction times associated with schizotypal disorders (Cannon, Byrne, Cassidy, Larkin, Horgan, & Sheppard, 1995; DeLisi, Svetina, Razi, Shields, Wellman, & Crow, 2002; Francks, Maegawa, Laurén, Abrahams, Velayos-Baeza, & Medland, et al., 2007; Gureje, 1988; Lenzenweger, 2001; Preti, Sardu, & Piga, 2007; Sommer, Ramsey, Kahn, Aleman, & Bouma, 2001) dyslexia (Rutter, 2004), autism (Boucher, 1977; Colby & Parkinson, 1977; Gualtieri & Hicks, 1985) and homosexuality (Lauman, Gagnon, Michael, & Michaels, 1994; Marchant-Haycox, McManus, & Wilson, 1991).

What is clear is that males are consistently more likely to demonstrate a greater degree of left-handedness than females (e.g. Annett, 1985, Bryden, 1977; Bryden & Roy, 2005; Chapman & Chapman, 1987; Gilbert & Wysocki, 1992; Lansky, Feinstein, & Peterson, 1988; Perelle & Ehrman, 1994, Oldfield, 1971; Reiss & Reiss, 1997). It is also clear that handedness has an affective role to play in language lateralisation (Khedr, Hamed, Said & Basahi, 2002; Kimura & Harshman, 1984; Knecht, Dräger, Deppe, Bobe, Lolmann, Flöel, et al., 2000). What is less clear, perhaps the essential kernel of the problematic, is an apparent aporia in a consistently reliable explanation for the degree of variability found between those mentioned here and the many other studies besides. Notwithstanding the coincidence of some researchers have not been able to show this handedness effect at all (e.g. Cornell & McManus, 1992; Green & Young, 2001; Holtzen, 1994; Salmaso & Longoni, 1985). Though, as Porac & Coren (1981) and Lansky, Feinstein, & Peterson (1988) have shown, non-significant results might be explained in not inconsiderable part by deficiencies in sample sizes and non-random sampling, respectively.

The situation of gender differences in handedness is perhaps far from straightforward at this time, as there appear to be three main overarching theoretical arguments in play in contemporary thought on the subject: the right-shift hypothesis (Annett, 2002); the modifier-gene hypothesis (McManus & Bryden, 1992); and, the recessive model (Jones & Martin, 2000). That said, however, each hypothesis diverges greatly from the others in terms of both structural mechanics and the degree of predictiveness available in accounting for variabilities in gender and handedness across the numerous studies available. In an attempt to combine the results from these three distinct hypotheses, some researchers have conducted painstaking meta-analyses of odds ratios in an attempt to reach a confluent basis for the quantitative evaluation of gender and handedness (e.g. Papadatou-Pastou, Martin, Munafò, & Jones, 2008; Rosenthal & DiMatteo, 2001; Silverman, 2006).

Notwithstanding, not all research into gender and handedness has led immediately to a greater elucidation of the subject. For example, Hepper, Shallidulah, & White (1991) have described laterality preference in foetuses of fifteen weeks old. Strong movement preferences (right over left arm) have also been shown in foetuses of barely ten weeks old (Hepper, McCartney, & Alyson, 1998). Moreover, follow up studies have confirmed that early thumb preference is an excellent predictor of handedness in later life (Hepper, Wells, & Lynch, 2005). Stirling & Elliott (2008: pp. 60-61) point up the bewildering implications of the Hepper, McCartney, & Alyson (1998) and Hepper, Wells, & Lynch (2005) studies, not least because, if true, “laterality preference would predate any overt indications of asymmetry in the developing brain.” A finding such as that surmised by Hepper, McCartney, & Alyson (1998), if true, would surely serve to muddy the waters of current debates into adult handedness and hemispheric specialisation still further.

Nonetheless, evolutionary psychologists such as Annett (1985) have espoused that, whilst approximately one in ten (10%) humans is left-handed, the degree of left or right-handedness is subject to variability. Furthermore, it is now reasonably well established that in studies looking at hemispheric differences (Johnson & Harvey, 1980; Nagae, 1985a, 1985b; Yen, 1975; Zaidel, 1987, 1988) handedness would appear to be a more reliable predictor of cerebral laterality, or indeed, cognitive processing ability, than is gender.

In the early days of neuro-imaging, Levy (1969) and Levy & Reid (1978) mapped differences in cerebral laterality between left- and right-handers, their findings showed that left-handedness led to less cerebral differentiation than is observable in right-handedness. Nagae (1985), in contradistinction to Levy (1969) and Levy & Reid (1978), suspected that the specialisation of language in the left-hemisphere had to be more evident in right-handers than in left-handers. As an adjunct to the previously mentioned findings by Levy (1969) and Levy & Reid (1978), Nagae (1985a, 1985b) claims to have found evidence to buttress the suggestion that, whilst right-handers store and process verbal and spatial information in separate cognitive systems, left-handers have comparative difficulties processing verbal and spatial information separately. Nagae (1985a, 1985b) would appear, therefore, to support evidence compatible with a suggestion that differences in cognitive processing revealed through handedness studies appear to be related to differences yielded because of the degree of hemispheric asymmetry between participants. Similar studies from research predating that of Nagae (1985a, 1985b), however, appears not to agree, only having reported negative effects in respect to a difference in asymmetric cognitive processing between handedness groups (Birkett, 1980; Hardyck, 1977; Newcombe & Radcliffe, 1973).

Other researchers appear to be more tempting of controversy with claims espousing that their findings show males and left-handers are both groups with an improved likelihood of displaying positive effects under test conditions when directly compared against their natural contraposition (e.g. Adam, Paas, Buekers, Wuyts, Spijkers & Wallmeyer, 1999; Bartélémy & Boulinquez, 2002, 2001; Boulinquez & Bartélémy, 2000; Bryden, 2002; Dane & Erzurumluoğlu, 2003; Der & Deary, 2006; Miller & Van Nes, 2007; Noble, Baker & Jones, 1964; Peters & Ivanoff, 1999; Welford, 1980). Moreover, in their meta-analysis of 144 studies in relation to the gender and handedness problematic, Papadatou-Pastou, Martin, Munafò, & Jones (2008: p. 677) concede that “the putative sex difference in handedness is, if reliable, one the most important constraints shaping our understanding of human handedness.”

Nevertheless, one might reasonably ask if this seemingly incongruent portrayal of gender difference is an accurate reflection of what is really being measured? Bryden (1977), for instance, has supposed that gender differences might be argued to be nothing more or less than a measurement artefact, maybe stemming from different reactions made by the respective genders to the hand preference tasks confronting them. There is also the important question of the roles played by uncertainty, obedience, and social conformity in individualist and collectivist cultures in handedness (Hofstede, 1980, 1983, 2001). Thankfully, long consigned to the historical record are reports of the so-called ‘gauche’ child forced to learn to use their right-hand for writing at school. Regrettably, it is more likely that contemporary social psychological research studies might instead consider relative success rates in females shifting from left to right-handedness due to differences in cultural values and attitudes in different international settings (Lansky, Feinstein, & Peterson, 1988; Porac, Coren & Searlman, 1986; Shimizu & Endo, 1983; Suar, Mendel, Misra, & Suman, 2007; Thompson & Marsh, 1976). Among those that have studied these variations in conformity to cultural norms, Hofstede (1980, 1983, 2001) can be said to have been hugely influential in the extensive classification of varying cultural attitudes enabling greater understanding of gender and handedness issues (see also Kirkman, Lowe, & Gibson, 2006). Never more so, perhaps, than in his now-famous definition (Hofstede, 2001: p. 9) of culture as “the collective programming of the mind that distinguishes the members of one group or category of people from another.”

Due to the sheer wealth of evidence for and against theories of gender and handedness relatedness emergent from empirical investigation and common experience alike, a consensus of reliability and validity measures in gender and handedness research can appear at times to be some way distant. Nonetheless, the formerly mentioned studies and others can provide a platform to form a notional acceptance of the scale of the subject. Moreover, in respect to the magnitude of the wider subject, the research strategy employed in this study confined itself to garnering data obtained from a small sample (N=40) within an adult age range of 18 to 50 years (see also Der & Deary, 2006; Jevas & Yan, 2001; Luchies, Schiffman, Richards, Thompson, Bazuin & DeYoung, 2002; Rose, Feldman, Jankowski, & Caro, 2002; Welford, 1977) for reasons that shall be explained later (see Age of participants).

One of the main attractions for a small sample approach to basic research into gender differences in handedness (i.e. hand preference) turns upon the perception of a lacuna of and for a practical understanding of the implications to be weighed from the many conflicting contemporary research claims (e.g. Der & Deary, 2005; Green & Young, 2001).

Age of participants

Test mean simple reaction time scores taken from outside the age range 18 and 50 years demonstrate roughly non-linear increases in reaction times (e.g. Der & Deary, 2005; Luchies, Schiffman, Richards, Thompson, Bazuin, & DeYoung, 2002; Verhaeghen & Salthouse, 1997). Reaction time test scores taken from <18 and >50yrs samples have often been of particular research interest in the context of evolutionary psychology  (e.g. Bleie, 2004; Kimura, 1973, 1992, 2002, 2004; Kimura & Hampson, 1994), developmental psychology (e.g. Hepper, McCartney & Alyson, 1998, Hepper, Shalidullah & White, 1991; Hepper, Wells & Lynch, 2005) and the psychological study of cognitive aging (e.g. Botwinick, 1966; Botwinick & Thompson, 1966; Gorus, De Raedt, Lambert, Lemper & Mets, 2008; Jevas & Yan, 2001; MacDonald, Nyberg, Sandblom, Fischer & Backman, 2008; Myerson, Robertson & Hale, 2007; Surwillo, 1973; Weiss, 1965; Welford, 1977). That having been said, increased age of participants has been shown to have an almost linear inverse effect on the number of instances of left-handedness across genders. That is to say, as age increases so too does the number of left-handers across genders decrease with almost inverse linearity (e.g. Annett, 1973; Ashton, 1982; Brackenbridge, 1981; Coren & Halpern, 1991; Dargeant-Paré, DeAgostini, Mesbah, & Dellatolas, 1992, Delattolas, Tubert, Catresana, Mesbah, Giallonardo, Lazaratou, et al., 1991; Gilbert & Wysocki, 1992; Lee-Feldstein & Harburg, 1982; Maehara, Negishi, Tsai, Otsuki, Suzuki, Takahashi, et al., 1988; McGee & Cozad, 1980; Porac & Coren, 1981; Salmaso & Longoni, 1985; Schacter, Ransil, & Geschwind, 1987; Smart, Jeffrey, & Richards, 1980).



The participants were a random sample of adults from University of Essex campus (N=22) and members of the public approached in Colchester Town centre (N=18). A total sample population size of forty (N = 40, M/F ratio = 1:1, LH/RH ratio = 1:1) participants took part in the outdoor study. The composition of participants was Males (N=20, 50%), Females (N=20, 50%), Left handed (N=20, 50%) and Right-handed (N=20, 50%). Following from suggested recommendations found in established reaction time testing literature (e.g. Luce, 1986; Sanders, 1998) each participant’s reaction times were tested a total of one-hundred times (100 tests x 40 participants) (i.e. 1S1R[1]=100, total reaction time trials = 4,000).

The effect of the participants’ age (years) is clearly of noteworthy interest, especially so to cognitive psychologists involved in the chronometry of gerontological research (e.g. Botwinick, 1966; Der & Deary, 2006; Gorus et al., 2008; Hultsch et al., 2002; Jevas & Jan, 2001; Lajoie & Gallagher, 2004; Luchies, Schiffman, Richards, Thompson, Bazuin, & DeYoung, 2002; Redfern et al., 2002; Rose, Feldman, Jankowski, & Caro, 2002; Welford, 1977). That being so, it was decided at an early stage in the research strategy that, whilst age is clearly a factor deserving further investigation, age data outside of the range 18-50yrs did not lie within the defined scope of the present study.


A laptop personal computer running Windows XP operating system and the ‘Psych/Lab for XP’ software package were preconfigured in anticipation of the participants’ readiness to participate in the experiment. The materials used in the present research were as follows: laptop personal computer (Toshiba Satellite U200-20U Part No. PSU44E-05F00JG3) running Windows XP service pack 3 (1); Psych/Lab for XP version 1.0 ‘Donders RT’ test software (1); MS Excel spreadsheet software (1); and, SPSS statistical analysis software. As a matter of course, all simple reaction timings were conducted on top of solid and static structures (e.g. brick walls) following the findings of Hsieh, Lin, & Chen (2007).


The statistical analyses for this study were composed of a single two-way analysis of variance (2×2 ANOVA, Gender * Handedness), and a single one-way analysis of variance (Age * Simple reaction time). A single dependent variable (DV) was the mean simple reaction time measured in milliseconds (msec) on a regular interval scale. The two independent variables (IV or factors) were Gender (M/F) and Handedness (L/R) of participants (each nominal scale).


Adults were approached on Essex campus grounds and in Colchester town centre. The researcher, holding a sealed empty glass jar, politely asked whether he might get some help opening the jar in exchange for a small reward. Upon completing the task, the helpers were given a small reward and asked politely if they had used their preferred hand to open the jar (after Annett, 1970). If the answer was affirmative, then the helper was asked whether they might agree to undergo a very short (i.e. approximately seven minutes) reaction time test. Again, only if the answer to the second question (i.e. taking part in the study) was positive were the helpers invited to continue to use their hand preference during the software test (the laptop computer was then placed on a previously selected solid flat surface (e.g.  a brick wall)). This initial part of the procedure served three functions: to establish the hand preference of the person; to introduce a degree of rapport and trust between the researcher and the participant; and, to generate the modest amount of interest necessary to recruit participants.

The procedure then switched to utilising a bespoke Windows-style application software (Psych/Lab for XP) using a simple black background and white text in a DOS-box display screen presented to the participants on the laptop computer. On the initial screen (Sc1) is displayed the message, “Press any key to continue.” This screen is followed by another screen (Sc2) displaying the message, “In this block of trials: Push the </> key as quickly as possible when the X appears in the RIGHT box. Press any key to begin.” After any key has been depressed the screen changes to another (Sc3) where a plus sign (+) is centred in the middle of the screen. At a distance of 5cm to the right of the plus sign a square yellow box (1cm x 1cm) is also clearly visible. Here, then, the participant begins to practice pressing the </> key as quickly as the X appears in the yellow box to the right of the screen a total of thirty times prior to the actual reaction time test (i.e. 30 practice trials).

Once the practice period of over, the actual test begins by showing screens Sc1, Sc2 and Sc3 again exactly as before. At the end of fifty (50) reaction time trials (block 1) the screen displays the following message, “End of block – Take a short break. Press any key to continue.” After a short break the participant returned to screen Sc2 and Sc3 for another fifty reaction time trials (block 2) by pressing any key.

Thus, the experiment ran in two blocks with a break in-between (with no fixed duration) for each participant (i.e. total 100 reaction time trials). The mean simple reaction times were then manually recorded from the mean of the total of the two test blocks into a spreadsheet (see Table 5 Appendix). The experiment may be characterised as a simple reaction time test where there is one stimulus [a square yellow box on the right of the screen] and one required response [pressing the </> key as quickly as an X appears in the yellow box].

In order the change the parameters of the reaction time test to suit the design required, when asked to “Press any key to begin”, depressing the <P> key will allow for the various parameters that determine the overall structure of the experiment and the timings for each trial event to be reconfigured. The parameters codes that can be modified along with the default values and the reconfigured values used in this study are listed below:

Code 1. Number of trials per test block: (default=50) (reconfigured value=50)

Code 2. Total number of blocks per session: (default=9) (reconfigured value=2)

Code 3. Number of practice blocks: (default=3) (reconfigured value=1)

Code 4. Number of trials per practice block: (default=10) (reconfigured value=30)

Code 5. Inter-trial interval (msec): (default=500) (reconfigured value=500)

Code 6. Duration of error tone (msec): (default=500) (reconfigured value=500)

Code 10. x location of stimuli (between 0 and 32000):

(default=16000) (reconfigured value=16000)

Code 11. y location of fixation point: (default=16000) (reconfigured value=16000)

Code 12. Colour of target X (white, red, green, blue, yellow): (default=1) (reconfigured value=1)

Code 13. Shortest foreperiod (msec): (default=700) (reconfigured value=700)

Code 14. Longest foreperiod (msec): (default=1100) (reconfigured value=1100)

Code 15. Minimum allowed reaction time (msec): (default=100) (reconfigured value=100)

Code 16. Distance between fixation and stimuli (video units):

(default=5000) (reconfigured value=5000)

Code 17. Width of cue box (video units): (default=1000) (reconfigured value=1000)

Code 18. Height of cue box (video units): (default=1300) (reconfigured value=1300)

Code 19. Colour of boxes (white, red, green, blue, yellow): (default=5) (reconfigured value=5)

Code 20. Maximum allowed reaction time (msec): (default=1500) (reconfigured value=1500)

Simple reaction times were collected from the standardised reaction time software (i.e. Psych/Lab for XP) running on a portable computer (i.e. Toshiba Satellite U200-20U) for forty participants. The results of the combined block test results (i.e. mean SRT) were then transferred into a Windows-style spreadsheet application (as the Psych/Lab for XP software is limited to storing just one participants’ data file at a time; see Limitations of the project) (see data on memory-stick or Table 5 Appendix). Finally, the data was analysed using statistical application software (i.e. SPSS v.16).


Mean simple reaction times

Mean results by gender appear to indicate that male mean simple reaction times were on average faster than female mean simple reaction times (Male = 257.55 msecs, Female = 275.85 msecs) (see Table 1 Appendix). Standard error of measurement for the same sample sizes, however, reveals less variability from the mean for males and a higher degree of variability between females (Female = +/- 8.76 msecs, Male = +/- 5.90). Furthermore, mean results by handedness appear to indicate that left-handers were on average faster than right-handers in their mean simple reaction times (LH = 252.65 msecs, RH = 280.75 msecs). Again, however, standard error of measurement for the same sample size and ratios of participants expose less variability (dispersion) from the mean for right-handers and only a marginally higher degree of variability between left-handers (LH = +/- 7.60 msecs, RH = +/- 6.48 msecs). Findings such as these would appear to prompt a proposal that, whilst left-handers were the fastest group in the study, males were the group with the lowest standard error of measurement. Further, taken together the mean results from this study might lead one to an interim proposal that, whilst males appear comparatively more consistent in their reaction times than females, left-handers of either gender might be faster in responding to eye-hand stimulus than right-handers, though left-handers might be adjudged to be more or less as consistent in their responsiveness over time when compared to right-handers. That said, at this stage an analysis of variance had not taken place.


Results from a two-way analysis of variance (2×2, gender (M/F) x handedness (LH/RH)) appear to be more revealing than the simple comparison of means analysis might have indicated (see Figure 1 below or Table 2 Appendix). Following the findings described by the comparison of means analysis, male mean simple reaction times are indeed faster than those compared to female scores (F(1,19) = 3.51, p = 0.069). Nonetheless, a between-participants analysis (gender x handedness) revealed that male left-handers were not only the fastest group, but also the group which recorded the lowest variability, that is, lowest mean simple reaction time and most consistent responsiveness (Mean = 245.50 msecs, SE = +/-5.54 msecs) (see Table 1 above). In contrast, the group with the highest variability in mean simple reaction times, and least consistent responsiveness, were female left-handers (Mean = 259.80 msecs, SE = +/- 14.20. msecs). Average differences in mean simple reaction time scores between female right-handers and female left-handers (Avg. diff = FRH 291.90 – FLH 259.80 = 32.1 msecs) were more pronounced than the difference in mean scores between male right-handers and male left-handers (Avg. diff = MRH 269.60 – MLH 245.50 = 24.1 msecs).

Figure 1 (above) showing mean simple reaction time between participants by gender

Figure 2 showing mean simple reaction time between participants by handedness

Handedness (Hand preference)

Again, the results of the two-way analysis of variance appear to show that, whilst there is a marginal difference in mean simple reaction time between genders, there was a significant and robust difference in mean reaction time (msecs) in favour of left-handers over right-handers (F(1,19) = 8.27, p = 0.007) (see Figure 2 above or Table 2 Appendix). This finding, if reliable, would appear compatible with the speculations made by other researchers that left-handers demonstrate significant positive differences in responsiveness compared to right-handers; typically using disproportionate success in professional careers and eye-hand sports as the meter of difference (e.g. Aggleton & Wood, 1990; Annett, 1999b; Azémar, & Stein, 1994; Cornell & McManus, 1992; Grouios, Tsorbatzoudis, Alexandris, & Brakoulis, 2000; Papadatou-Pastou, Martin, Munafò, & Jones, 2008; Peters & Ivanoff, 1999; Peterson, 1979; Searleman, Herrmann, & Coventry, 1984; Wood & Aggleton, 1989, 1991).

The two-way analysis of variance also showed a non-significant interaction of main effects (gender * handedness) (F(1,39) = 0.168, p = 0.685) (see Table 2 Appendix). However, average differences in mean simple reaction time scores between left-handers and right-handers of either gender indicate that left-handers were evidentially more consistently responsive compared to right-handers (Avg. diff = FLH 259.80 – MLH 245.50 = 14.3 msecs, FRH 291.90 – MRH 269.60 = 22.3 msecs). Right-hander mean simple reaction time scores irrespective of gender appear to indicate a large increase (156%) compared against left-handed score variability.


It is clearly important that age did not have an affective role to play in the results garnered in this study. In this respect, a one-way analysis of variance (mean simple reaction time by age) reveals a non-significant effect with respect to age and mean simple reaction time (F(1,16) = 1.182, p = 0.349) (see Table 3 Appendix). This result would appear to therefore point to age of participants as not being a factor in the results gained in this study. This non-significant finding for age might further be noted as adding support to the claims made by other researchers that the adult age range of 18-50yrs can be said to be notionally representative of non-child reaction time scores (e.g. Der & Deary,  2005; Luchies, Schiffman, Richards, Thompson, Bazuin & DeYoung, 2002; Verhaeghen & Salthouse, 1997). Although a non-significant finding, it also appears to be clear that the majority of those preferring to use their right hand come from the 22 – 27yrs age range, whereas those preferring to use their left hand come from a broader spectrum (22 – 50yrs) of ages. This handedness concentration can be accounted for through the wider search in the local community for adult left-handed participants beyond those found on Essex campus.

Figure 3 (Top Left) showing mean simple reaction time scores between participants by gender and age

Figure 5 (Bottom Left) showing standard deviation of reaction time scores between participants by gender and age (interpolated for missing age values)

Figure 4 (Above Right) showing within participants mean simple reaction time

by gender and age (interpolated for missing age values)


Gender Differences

Evidence obtaining from this study shows that mean simple reaction times approached significance for males compared to females (F(1,19) = 3.51, p = 0.069). This finding appears to be at odds, with those findings from other studies placing emphasis on a significant and strong effect for gender differences (e.g. Adam, Paas, Buekers, Wuyts, Spijkers, & Wallmeyer, 1999; Dane & Erzurumluoğlu, 2003; Der & Deary, 2005; Noble, Baker & Jones, 1964; Silverman, 2006; Welford, 1980). Nevertheless, evidence from the current study appears to show that, whilst there was an approaching significance difference in mean simple reaction times between genders, the results were non-trivial. As such, Type I errors cannot be excluded, that is to say, a larger sample population size might be suggested to investigate further whether differences between genders might be more clearly demonstrated between males and females dependent on mean simple reaction time as a basis of understanding. Although, as previously mentioned, it is perhaps noteworthy that some much larger studies have not been able to find evidence to support a suggestion for gender differences. That said, for Der & Deary (2006: p. 71) a not improbable explanation for gender differences in reaction time studies might be “in the speed-accuracy trade-off.” However, any relation said to exist between gender and number of errors may only be speculated upon, as this study’s results are based upon correct responses only.

In contrast, evidence collected in a meta-analysis of gender differences in handedness studies by Silverman (2006) appears to show that difference in visual reaction time performance between genders might be gradually decreasing over the last few decades. Silverman (2006) attributes these findings in part to increased equality, as well as, in larger part, attributing gradual descreases in visual reaction time performance to the actuality that more women are participating in historically male pastimes (e.g. high-impact sports).

That said, the finding from this research appear to be compatible with a speculation that adult males might be consistently more responsive in simple reaction time tests compared to adult females. Even if this finding was indicative of a pattern, however, the possibility cannot be discounted that similar or dissimilar samples might show different results on day by day, or even a moment by moment testing periodicity; although, the variability of individual differences has received a considered and measured treatment by Rabbitt, Osman, Moore, & Stollery (2001). That being so, variability in gender effects do still regularly appear within large numbers of studies, and the seat of controversy – variation and incongruity between findings – has been variously attributed by researchers to a raft of factors over time ranging from individual differences (e.g. Garavan, Hester, Murphy, Fassbender, & Kelly, 2006; Hardyck, 1977; Martin & Jones, 1999a; Nettelbeck, 1973; Overby, 1994; Rabbitt, Osman, Moore & Stollery, 2001), endocrinological factors (e.g. Kimura & Hampson, 1994) and the nature and degree of cerebral hemispheric specialisation or laterisation (e.g. Annett, 1985; Coren, 1995; Crow, Crow, Done, & Leask, 1998; Genetta-Wadley & Swirsky-Sacchetti, 1990; Hardyck, C. (1977; Hardyck & Petrinovich, 1977; Knecht, Dräger, Deppe, Bobe, Lolmann, Flöel, et al., 2000; Lake & Bryden, 1976; Miller & Van Nes, 2007; Nagae, 1994; Previc, 1991; Rassmussen & Milner, 1977; Shan-Ming, Flor-Henry, Dayi, Tiangi, Shuguang, & Zenxiang, 1985; Smith, 1938; Springer & Deutsch, 1998; Zaidel, 1985).

Patterns of Handedness

The most salient finding admitting to a pattern of handedness was the significant result strongly favouring left-handers over right-handers (F(1,19) = 8.27, p = 0.007). Moreover the faster and most consistently responsive group of participants were the left-handed males (Mean = 245.50 msecs, SE = +/-5.54 msecs) (see Table 1 Appendix). In addition, the average difference between left-handed and right-handed mean simple reaction times, irrespective of gender, showed an average difference in variability (M/FLH = 14.3 msecs, M/FRH = 22.3 msecs) favouring left-handers. Such a finding, if reliable, alongside the handedness effect shown above in a two-way analysis of variance, would appear to point to a presumption of a left-handed advantage in simple reaction time tasks. Unfortunately, whilst not the slowest group (Female RH = 291.90 mean SRT msecs), the least consistent group were female left-handers (Mean = 291.90 msecs, SD 44.90). When viewed together as a whole, therefore, it would appear that one can make a reasonable claim for the reaction speed of left-handers over right-handers, however, one cannot also claim that the faster reaction speeds shown by left-handers might also be matched by their prowess in the equally important area of consistency. Given the moderately inconsistent nature of left-hander reaction speed remaining irreconcilable with a clearly demonstrable finding of reliability between participants in this study, further re-testing using similar sample sizes may be warranted to yield more definitive results (see Reaction Time Variability).

Reaction Time Variability

The variability of reaction time scores in this study showed that, whilst males with a left hand preference had the lowest variability of scores from the mean (SD = 17.53), females with a left hand preference had the highest variability of scores (SD = 44.90). This finding might perhaps be explained as a consequence of anomalous results from two particular female left–handed participants (Nos.16 & 17, see Table 5 Appendix) both aged 46yrs whose scores were the most variable of the sample population (SD = 66.47, SE = +/-47.00). It is noteworthy that without the inclusion of these two deviants (Nos. 16 & 17) it is unlikely that female left-handers variability would have been quite so affectively dispersed.

In this study the mean simple reaction time correlation coefficient of variation is shown to be non-linear and not indicative of any remarkable gender difference. Moreover, it is perhaps more remarkable that this finding was once again found to contain uncertainty, indefinites, an equivocal situation no less. For, as previously mentioned, variability in gender and handedness differences have been widely attributed to be multifactorial, inclusive of but not limited to individual differences (e.g. Garavan, Hester, Murphy, Fassbender, & Kelly, 2006), endocrinological factors (e.g. Kimura & Hampson, 1994), and, cerebral hemispheric specialisation or cerebral laterisation (e.g. Annett, 1985), to name just three examples. What is clear is that continued studies into the area of gender and handedness variability surely have much to hold researchers’ ongoing interest. Not least in the cognitive psychological disciplines, but also perhaps in terms which are inclusive of mixed-model multi-disciplinary studies which may gain further purchase from the overlap into the fields of individualism, conformity, belief and educational attitudes (see also Hofstede, 1980, 1983, 2001).

Strengths of the study

The main strengths of the project are perhaps twofold. First, there is the quality of the sample, as it was garnered from over 4,000 separate reaction time tests (i.e. 100 reaction time tests x 40 participants). Second, the Psych/Lab for XP software application ‘Donders RT’ test used is of a standard format, is easy to configure to suit requirements, and uses a memory footprint sufficiently small (executable file size 118Kb) allowing for easy installation and reconfiguration where necessary.

Limitations of the study

One main limitation of the project is that the software application (i.e. Psych/Lab for XP 1.0) used to measure reaction time did not store the results of the individual trials. Notwithstanding, the mean simple reaction times for all participants were manually recorded on a spreadsheet prior to further analyses (see attached memory-stick or Table 5 Appendix). Another main limitation of this study is that results were based solely on correct responses. Future research on this topic might well benefit from an analysis of the coefficient of error variability. This design issue (i.e. the use of error as a factor) thus precluded this study from any further speculations concerning potential speed versus accuracy measures. That having been noted, the matter of using the number of errors by gender and handedness as a factor was considered at the outset of the research strategy to be an added complication which might detract from the basic concern of the study – that is, responsiveness. The matter of this study being conducted outside the laboratory, in actuality in public spaces, does certainly allow for the possibility of criticisms regarding less rigorous conditions, and thus greater variability, than those found in laboratory experiments. However, given the rehearsed nature and rigidity of the set procedure followed, every attempt was made by the researcher to maintain a consistent and formal outdoor experiment.

Potential concerns might also be levelled at the degree of keyboard latency timing for the particular model of personal computer used in the testing section of this study. In answer to this concern, a single laptop computer (i.e. Toshiba Satellite U400-20U Part No. PSU44E-05F00JG3) ran a single application during the reaction time study (i.e. to reduce potential hardware or software processing latency). More specifically, during transmission from keyboard to system, the keyboard used in this study checks the clock line for low-level transmission at intervals ≤ 60 microseconds (µsecs) and the data line interval at ≤ 10 milliseconds.[2] From these timing measures, it is clear that keyboard to clock line cannot be a significant factor in this study as ≤ 60 microseconds is equivalent to ≤ 0.06 msecs. However, a keyboard to data line interval timing of ≤ 10 msecs could present certain difficulties unless all participants had used the same keyboard. Indeed, it is certainly noteworthy to repeat that the same computer and keyboard ought to be used throughout these types of studies with all participants.


The present study provides a descriptive investigation into adult gender and handedness (i.e. hand preference) in simple reaction time responsiveness from a symmetrical population sample (i.e. M/F = 1:1, L/R = 1:1). Here different patterns of responsiveness results were found for gender and hand preference. In this study the relation between gender and responsiveness approached significance, but was not yielding of any definitive pattern, thus leaving gender and responsiveness an equivocal concern certainly warranting further research. In contradistinction to the findings shown for gender and responsiveness, this study showed a significant and strong relation between hand preference usage and responsiveness in favour of left-handedness compared to right-handedness. Indeed, the positive effect of hand preference and responsiveness can be said to add a modest degree of further support to studies seeking to constrain the range of acceptable quantitative predictions of gender and handedness relations (e.g. Der & Deary, 2006; Papadatou-Pastou, Martin, Munafò, & Jones, 2008).

That said, at a more general level of observation one of the main subjects of this present study – handedness and hand preference asymmetry – might for some appear at first to be two different phenomena altogether (e.g. Bishop, 1989; Morgan & Corballis, 1978). It is therefore worthwhile noting the implications of a study focusing on asymmetrical hand preference differentials in those without asymmetrical skills (McManus, Murray, Doyle, & Baron-Cohen, 1992). In their study of what might appear to be hand preference seemingly without reason for such a preference, McManus, Murray, Doyle, & Baron-Cohen (1992) successfully found evidence for hand preference within a population of young people considered to be on the autistic spectrum. McManus, Murray, Doyle, & Baron-Cohen (1992) appear to also underpin findings for asymmetric hand preferences in the absence of asymmetric skills in people with schizotypal disorders (DeLisi, Svetina, Razi, Shields, Wellman, & Crow, 2002). Following on from the findings of Dellatolas, Pascale, & Curt (1997), and McManus (1983), each comes to an evidential basis for concluding the correctness of dissimilar interpretations when drawing a distinction between the degrees of handedness and the nature of laterality (or directionality). That is, Dellatolas, Pascale, & Curt (1997) and McManus (1983) explicitly support the correctness of drawing demarcations between degrees of handedness and the nature of laterality/direction of that handedness, whilst also implicitly freighting support for the equipollence of the notional terms handedness and hand preference asymmetry.

Continuing on the important theme of the nature of degree of handedness and laterality/direction, it is noteworthy that Dassonville, Zhu, Ugurbil, Kim, & Ashe (1997) used their findings taken from functional magnetic resonance imagery (fMRI) to propose that the degree of handedness and laterality/direction are located in discrete areas of the cerebral cortex. Neuroimaging work by Habib, Gayraud, Oliva, Regis, Salamon, & Khalil (1991), and Witelson (1989), appear to agree to having found increased hemispheric bandwidth differences in the ambidextrous and left-handers compared to right-handers affected by gender. That is to say, a much greater degree of synaptic connectivity brought about by much larger corpus callosum tending to be shown in ambidextrous and left-handers alike, modulated by gender. Although, it is noteworthy that more recent neuroimaging evidence by Luders, Rex, Narr, Woods, Jancke, Thompson, et al., (2003) have been used as the foundation for speculations on the nature of the coextensive roles played by callosal intracerebral connectivity and sulci asymmetry. That being so, in many studies – the present study included – no attempt has been made to differentiate between the degree of handedness and laterality/direction in no small part due to economic considerations.

But what of the role of acculturation in gender and handedness responsiveness? Moreover, why is it that the core Scandinavian nations (e.g. Sweden, Norway and Denmark) do not appear to follow the findings which predict an increased likelihood of left-handedness in males? (See Papadatou-Pastou, Martin, Munafò, & Jones, 2008.) To these questions one might turn to the theories of cognitive social psychologists, such as those put forward by Hofstede (1980, 1983, 2001). For Hofstede (1980, 1983, 2001) one becomes better able to comprehend cross-cultural issues (here, for instance, gender and handedness relations), through attitudes, values, conformity, and uncertainty avoidance, as well as, the continuums of masculinity-femininity and individualism -collectivism. In his seminal work of cross-cultural psychology, Hofstede (1980) had introduced a paradigm by which traditional dichotomies between East-West, for instance, could be accurately analysed. Moreover, it is Hofstede’s (1980) paradigm which has proved so fruitful in the disconfounding of cross-cultural gender differences in handedness studies as in some part being modulated by cultural values and influences (Kirkman, Lowe, & Gibson, 2006; Schimmack, Oishi, & Diener, 2005). In addition, one might also utilise Hofstede’s (1980) masculinity-femininity / collectivism-individualism continuums profitably with regard to the Scandinavian problem. Using Hofstede’s (1980) continuums in their meta-analyses of 144 gender differences in handedness studies, Papadatou-Pastou, Martin, Munafò, & Jones (2008: p 691) have recently reported that, whilst cultures manifesting the lowest levels of gender differentiation (i.e. reduced cultural masculinity) would also appear to be those evidencing the least measures of gender differences in handedness, those cultures with significant gender differences in handedness correlated with those with significant measures of uncertainty avoidance and significant diminutions alongside related to escalations in individualism. These findings, if they are reliable, are perhaps the first of this kind which will no doubt aid researchers considerably in their approach of the problematic explanation of cross-cultural gender differences in handedness. However, this suggestion does not mean to infer there is little or no difficulty in disconfounding the results from other possible social influences such as values, attitudes or compliance issues (see also Gabriel & Gardner, 1999; Van Vugt, De Cremer, & Janssen, 2007).

In summation of the present basic research, the study provided an objective test of the hypothesis that masculinity and/or left-handedness might have effects on adult eye-hand visual reaction times over and above those explained by individual differences. This study did not find evidence compatible with any definitive speculation regarding a natural superiority of one gender’s responsiveness over another, instead, perhaps, finding in favour for a more subtle interpretation inclusive toward genetic biology and the facilitating environment as an inseparable yet mutually discrete antinomy. The evidence shown in this study regarding handedness (hand preference) responsiveness is perhaps suggestive of (a) a significant innate tendency in laterality which remains ripe for further inter-disciplinary meta-analyses, and (b) a considered revisiting of the significance which ought to be placed on the utility of controlling for handedness in subsequent eye-hand visual reaction time studies.

[1] 1S = One given stimulus, 1R = One possible response

[2] As per manufacturers guidelines (see references section Toshiba, 2008)


2 thoughts on “Adult Gender and Hand Preference Differences in Eye-Hand Visual Reaction Times

  1. Thanks so much Rahil. Much appreciated. Reference (Wadey, P. (2009). Adult Gender and Hand Preference Differences in Eye-Hand Visual Reaction Times. Wivenhoe: Essex University, unpublished)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s