Abstract
This study compared right-hand and left-hand two-choice CRT responses of college students to both the simplest chronoscope (a falling meterstick) and to a digital-readout multi-choice reaction timer (machine). Tests of ipsilateral differences showed significant differences in CRT between the meterstick and the machine methods for both right and left hands. There were no significant contralateral differences in CRT with either the meterstick or machine methods. Combined total scores showed that there were significant differences in CRT responses obtained by meterstick versus machine, whereas there were no differences in CRT responses for all right-hand responses versus all left-hand responses. Results suggest that perhaps the workings of two different reaction time systems may account for the robustness of differences between CRT's obtained by meterstick versus machine.
This study is the fourth of a series of studies describing the empirical results obtained by the use of the simplest chronoscope (a free-falling meterstick) to measure visual reaction time that once again employed a within-subject, repeated-measures design to attempt to reduce variability because each participant served as his/her own control. In the first study (Montare, 2009), visual simple reaction time (SRT) obtained by use of the simplest chronoscope (meterstick RT) was found to be significantly faster than long-established population approximations of SRT. In the second study (Montare, 2010) in a comparison between SRT obtained by meterstick to that obtained by use of a typical digit-readout, multi-choice reaction timer (machine RT), it was found that meterstick RT was significantly faster than the long-established population mean (200 msec.), whereas for the same individuals machine RT was significantly slower than the population mean. In the third study (Montare, 2013), meterstick RT was compared to machine RT for all three of Donders' (1868/1969) methodologies: (a) meterstick SRT was significantly faster (Cohen's d = 2.85) than machine SRT; (b) meterstick choice reaction time (CRT) was significantly faster (Cohen's d = 2.59) than machine CRT; and (c) meterstick discriminative reaction time (DRT) was significantly faster (Cohen's d = 2.80) than machine DRT.
The most salient empirical conclusion to be drawn from the previous studies summarized above is that in all prior comparisons between meterstick RT and either long-established population approximations of RT or observed scores from the use of a modern digital reaction time methodology (machine RT), responses were significantly faster when obtained by the simplest chronoscope. The magnitude of the above Cohen's d values suggest that these empirical differences between the meterstick and machine methods might require a theoretical account for what may well be two different reaction time systems for processing signals in different visual RT tasks.
The purpose of the present study was to extend previous investigations to encompass the empirical study of right- and left-hand differences in two-choice CRT scores obtained by the meterstick method to those obtained by the machine method. In the present study, the same within-subject repeated-measures design was used as in Montare (2009, 2010, 2013) to reduce the influence of the ubiquitous individual differences that have been noted in RT studies since the time of Helmholtz (as cited in Woodworth, 1938) who abandoned the method in the 19th century because he observed such high variance in responses. Here, an assessment of variability in choice reaction time obtained by meterstick versus machine will address the long-standing problem of high variability in reaction time responses. The CRT design common to both methods called for the presentation to college students of two visual imperative stimuli with instructions to quickly respond to imperative Stimulus 1 with a right-hand response and to quickly respond to imperative Stimulus 2 with a left-hand response (complete details of the procedure may be found in Montare, 2013).
Ipsilateral Differences
Hypotheses 1 and 2 of the present study extended our previous investigations to now include tests of the ipsilateral responses obtained by meterstick versus machine. For the first time this study will examine same-hand differences between the two different methodologies based on the specific prior finding in Montare (2013) noted above, that overall CRT scores obtained by meterstick were significantly faster. Tests of Hypotheses 1 and 2 below will provide evidence about the extent to which the significant differences between overall CRT scores would also be observed when each of the two hands were separately examined.
Hypothesis 1. If the observed significant differences between overall meterstick CRT and machine CRT can also be demonstrated to exist between the right-hand ipsilateral responses to the two different methodologies, then right-hand responses by the same subjects to meterstick CRT should be significantly faster than their ipsilateral right-hand responses to machine CRT.
Hypothesis 2. Left-hand responses by the same participants to meterstick CRT should be significantly faster than their ipsilateral left-hand responses to machine CRT.
Contralateral Differences
In a review of the early literature on contralateral reaction times between the two hands, Woodworth (1938) concluded that“…it makes little difference which hand is used…” (p. 329) because the average difference in previous studies between the hands was only about 5 msec. More recently, Nisiyama and Ribeiro-do-Valle (2014) also found no differences between right-hand and left-hand responses to CRT. Seashore and Seashore (1941) showed that average auditory reaction times were 147 msec. for the right hand and 144 msec. for the left hand; the.92 correlation between the RTs suggested that the two hands respond using common underlying processes. Additional evidence of equal right- and left- hand responses provided by Benton and Joynt (1959) showed approximately equal reaction times in both hands of control participants in a two-choice reaction time task. Thus, within the same two-choice CRT task there should be no significant contralateral difference in reaction time observed. However, contradictory evidence exists. Some studies have found that right-hand CRT is faster than left-hand CRT (Kerr, Mingay & Elithorn, 1963; Rabbitt, 1978); while at least one other study found that right-hand CRT was slower than left-hand CRT (Arnett & Arnett, 1979). Therefore, to provide further evidence:
Hypothesis 3. There will be differences between the hands when CRT is obtained by the Montare meterstick method.
Hypothesis 4. There will be differences between the two hands when CRT is obtained by the standard machine method.
Overall Total Differences
Hypothesis 5. When total scores obtained for CRT reactions to both right and left hands are combined, there will be differences between meterstick responses compared to machine method responses.
Hypothesis 6. When total scores obtained for CRT reactions to right and left hands are examined, there will be differences between right-hand and left-hand RTs.
Method
Participants
The participants were 33 right-handed students of both sexes who attended William Paterson University (24 women: M age = 21.5 yr., SD = 2.1; 9 men: M age = 21.4 yr., SD = 1.2). All participants were undergraduate psychology majors enrolled in experimental psychology courses who volunteered to participate for course credit. Statistical analyses of the sex differences in response times indicated that, although there was a consistent tendency for males to have faster RTs than females, the observed sex differences between the means were not significant in any given comparison between right- and left-hand responses for either the meterstick or the machine methods. In addition, as expected from previous reports (Woodworth, 1938) Arnett & Arnett, 1979), there were no significant effects attributable to either sex or hand preference.
Apparatus
A hardwood meterstick graduated to the nearest millimeter was employed as the simplest chronoscope to obtain the meterstick CRT responses. For the machine method, a multi-choice reaction timer (Model 64013, Lafayette Instrument Company, Lafayette, IN) was used to obtain the machine CRT responses. A small electronic metronome was used to provide the 1 Hz clicks that timed the cue durations between the “ready signal” and the presentation of the stimuli used in the meterstick CRT procedure.
Procedure
In both meterstick- and machine-measured tasks, the participants were individually administered 200 trials (100 meterstick CRT trials; 100 machine CRT trials). Given that the procedures common to both the meterstick and the machine tasks are the same as in previous papers, complete details have been previously fully described in Montare (2009, 2010, 2013). The procedure for measuring CRT by the meterstick method was conducted by the use of a falling meterstick graduated to the nearest millimeter and was fully described in previous work (Montare, 2009. 2013). The procedure for measuring machine CRT by use of the Lafayette Multi-Choice Reaction Timer was carried out using the control panel to set the stimulus conditions for every trial and the response keypad to measure the reaction time that was digitally displayed to the nearest millisecond on the control panel. Once again, this procedure was fully described in Montare (2009, 2013).
Results and Discussion
The major purpose of the present study was to provide empirical tests of six hypotheses about the lateral differences in choice reaction time responses obtained from the same individuals by meterstick versus machine methods. Table 1 shows the means and standard deviations based upon a total of 200 trials per student (50 meterstick right-hand responses, 50 meterstick left-hand responses, 50 machine right-hand responses, 50 machine left-hand responses).
Descriptive Statistics Used to Test the Six Hypotheses
Response Variability and Differences
The data in Table 1 showing the standard deviations can be used to address the problem of variability by calculating standard Fmax tests between variances obtained from meterstick versus machine methods. The variability of the right-hand CRT scores obtained by meterstick was significantly less than the right-hand CRT score variability obtained by machine (F32, 32=2.05, p <.05), very similar to the finding for the variability of the left-hand CRT scores obtained the same ways (F32, 32=2.09, p <.05). The third result was that the variability of the total meterstick CRT scores was significantly less than the variability of the total machine scores (F32, 32 = 2.00, p <.05). Therefore, it may be concluded that the use of the simplest chronoscope significantly reduces the variability of CRT scores when compared to the typical variances obtained by machine methods. This reduction in variability could be used to address the problem of high variability noted by researchers in the general area of reaction time since Helmholtz, as noted in the Introduction of the present paper.
Table 2 displays the statistical tests that were performed to test each of the six hypotheses based on the results shown in Table 1. As may be seen in Table 2, the results fall into three categories: ipsilateral differences between responses by the same hands; contralateral differences between the hands; and, overall total differences when the data from the two hands are combined.
Summary if Tests of Six Hypotheses
p <.01 (two-tailed).
Ipsilateral Differences in Responses
Hypothesis 1 was that an ipsilateral comparison of the right-hand responses by the same participants in the meterstick CRT should be significantly faster than their right-hand responses to machine CRT. Table 2 shows that the mean lateral responses of the right hand to the meterstick method were 134 msec. faster (Cohen's d = 2.54). The magnitude of this difference strongly suggests that the two distributions could have been generated by two different reaction time systems. Furthermore, the correlation between the RTs was not significant (r =−.04), suggesting that the right hand responds to the two methodologies using non-correlated, separate, and independent central nervous system processes in producing the CRT responses.
Hypothesis 2 was an ipsilateral comparison of left-hand responses by the same participants to meterstick CRT, expected to be significantly faster than their left-hand responses to machine CRT. In Table 2 the mean lateral CRT responses of the left hand to the meterstick method were 146 msec. faster (Cohen's d = 2.67). The correlation was not significant (r =−.12), a further indication that the left hand shows the same pattern of ipsilateral responses to the two methodologies as does the right hand.
Based upon these results, it may be concluded empirically that in a two-choice reaction time task the same individuals responded ipsilaterally significantly faster with either hand to the simplest chronoscope than they did to a standard reaction time machine. Therefore, CRTs obtained from each hand separately are significantly faster when measured by meterstick than when measured by machine. The significant ipsilateral differences between the two methods indicates that perhaps two response systems may be responsible.
Contralateral Differences in Responses
Hypothesis 3 states the expectation that a contralateral comparison of right-hand responses by the participants to meterstick CRT with their left hand will differ from their responses to meterstick CRT. The mean lateral responses of the left hand to the meterstick method were only 3 msec. faster than the right-hand responses (Cohen's d = 0.22). Furthermore, the correlation between the 33 individual means of the two sampling distributions was significant and very strong (r31 =.90, p <.01). There are several conclusions: (a) there are no significant contralateral differences between the hands; (b) small differences have small effect sizes; and (c) the strong, significant correlations suggest that the two hands seem to be using similar underlying processes to produce CRT responses within each of the two methods.
Hypothesis 4 compared right-hand responses in machine CRT to the same participants' left-hand responses to machine CRT. The mean lateral responses of the right hand to the machine method are only 9 msec. faster (Cohen's d = 0.33). The correlation between the 33 individual means was large and significant (r31 =.79, p <.01). The left-hand results for contralateral comparisons display the same pattern of findings as the right hand.
Combining the results of Hypotheses 3 and 4, it may be observed that the non-significant contralateral differences in the present study are findings that are congruent with the long-standing observation previously cited above that“…it makes little difference which hand is used.” (Woodworth, 1938, p. 329) because the average difference between the hands is only about 5 msec.”
Overall Total Differences in Responses
It should be noted that the most salient findings contained in Table 2 were that the two-choice CRT means for responses to the meterstick method at 215 msec. were similar to those usually reported in the earlier literature; whereas the two-choice CRT means for the responses of the same individuals to the machine method (356 msec.) were similar to those usually reported in the current literature: in the present study the difference between the total scores obtained by the meterstick method were 141 msec. faster than the total scores obtained by the machine; this disparity between the mean responses to the two methods by the same individuals on the CRT task is a major empirical finding of the present study.
Testing Hypothesis 5, the observed difference of −141 msec. between the means was significant with a very large effect size (Cohen's d=2.74). The correlation between the overall combined total RTs to the meterstick compared to the overall combined RTs to the machine was not significant (r31 = -.07). This lack of significant correlation between the responses of the same individuals to the two methods, combined with the fact that the two means are significantly different, suggests that different processes are used to meet the demands of the two tasks. In addition, it should be noted that the variability of scores from the meterstick method was significantly smaller than the variability of scores of the machine method.
It was noted in the introduction that in all previous comparisons in our laboratory between the meterstick and the machine responses those obtained by use of the meterstick were robustly faster than the responses, by the same individuals, to the machine methodology. The present data extend this overall empirical finding to lateral differences, such that responses of each hand are also significantly faster when assessed by meterstick versus machine. The proposed hypothesis of two reaction time systems within the same individuals will be further investigated in our next paper based upon an expanded new database to further test differences between meterstick and machine.
The test of Hypothesis 6 examined differences of the responses between the hands when the RTs were combined for each of the two methods, i.e., the overall differences between the two hands. The responses for the two hands were only 3 msec. different, which was not significant; however, the correlation between the overall combined RT's of the right hand compared to the those of the left hand to both methods combined was significant (r =.85, p <.01). This significant correlation and the non-significant difference in the overall means suggests that the two hands are invoking similar internal processes to meet the demands of the reaction time tasks.
Finally, the differences in CRT responses obtained by meterstick versus machine in the present study should be integrated with the findings and the theoretical issues of the three other papers in this series of studies (Montare, 2009, 2010, 2013). This task shall be the subject of the next paper in this series, based upon an expanded new database to further study the suggestion of two reaction time systems.