Gender Differences in Stroop Effect

Gender differences were examined in the speed of processing, in the context of a Stroop colour-word task. Overall, the Stroop interference effect was observed. Word-reading was faster than colour-naming. Response times were much slower when naming the colour of incongruent colour words than series of Xs. Female participants seemed to respond faster in naming colours of incongruent colour words than males. In contrast, female and male participants performed equally in the word-reading and colour-naming tasks. The findings are insufficient to substantiate gender difference in the Stroop interference effect.

Keywords: Stroop interference, speed of processing, gender, incongruent colour words

Gender Differences in the Stroop Colour -Word Interference Effect

The Stroop color-word test typically consists of a basic task that acts as a control, either word-reading or colour-naming task; and an interference task that requires the participant to suppress the habitual response and provide an unusual one, e.g., naming the ink of incongruous colour words. In the latter task, colour naming takes longer and is more error-prone than a congruent colour word.

The Stroop interference effect (henceforth Stroop effect) refers to the latency to name colour words. The Stroop effect have been large and statistically significant in research studies, but not adequately explained. Two theories have attempted to explain the effect. The speed of processing theory states interference occurs as words are read faster than colours. The selective attention theory states naming colours requires more attention than reading words.

As a measure of processing speed, the Stroop test is widely used for research in selective attention, automaticity, inhibitory processes, and executive control (for review, see Dyer, 1973; MacLeod, 1991). The robustness of the test has earned its name as the “gold standard” of attentional measures (MacLeod, 1991, 1992). The stroop test is also used as a diagnostic instrument in clinical settings and research studies to help identify individuals with brain damage (Golden, 1976).

Performance on many neuropsychological tests, including Stroop, is influenced by demographic variables such as age, sex, and education (Mitrushina, Boone, & D’Elia, 1999, cited in Roivainen, 2010). Longman, Saklofske, and Fung (2007, cited in Roivainen, 2010) found the mean gender difference in the Stroop test was equivalent to two to four years difference in the years of education.

The causal relationship between rapid naming (RAN) skills, processing speed, and reading fluency has been an important issue in recent research on dyslexia in children (Puolakanaho et al., 2008). Knowledge of gender differences in the Stroop effect would help in identifying learning impairments in children in early education.

Gender Difference

The extensive literature regarding gender differences in the Stroop Test has been inconclusive (for review, see MacLeod, 1991; Roivainen, 2010). For every study that reported significant gender difference, there seemed to be one that claimed otherwise. Even so, there is a tendency for females to perform superior to male.

There is general consensus that females tend to have shorter latency for colour-naming (e.g. Golden, 1974; Sarmany, 1977; Strickland, Elia, James, & Stein, 1997, all cited in Alansari & Baroun, 2004). However, views on the effects of gender in word-reading and colour-word naming contrasted.

Armengol (2002, cited in Alansari & Baroun, 2004) administered the Stroop test in Spanish to 349 children, (174 boys, 175 girls), age 6 to 12 years, from two Mexican schools and reported significant gender differences. In contrast, Insua (2002, cited in Alansari & Baroun, 2004) did not find such difference among 2000 English and Spanish participants in the United States across all Stroop tasks.

Alansari and Baroun (2004) investigated performance differences related to gender and culture using the Stroop Color and Word Test for a sample of 140 Kuwaiti students (60 men, 80 women) and 70 British students (36 men, 34 women). They found significant Stroop interference with culture but not gender.

In contrast, Van der Elst, Van Boxtel, Van Breukelen, and Jolles (2006) found women outperformed men on Stroop interference and colour-naming tasks in a sample of 1856 Dutch participants. Age and education also had significant effects on all three Stroop trials. This finding is supported by Baroun and Alansari (2006) in a study with 504 Kuwaiti university students.

Aim of Experiment

This experiment is based on a modified version of Golden’s (1975, cited in MacLeod, 1991) Stroop colour-and-word Test. The Word (W) and Colour (C) controls were used to establish word-reading and colour-naming baselines respectively, while the Colour-Word test (CW) was used to measure colour-word interference.

For control trials, the response times for word reading (W) were expected to be faster than colour naming (C). For the test trial, the response times for colour-word naming (CW) were expected to be greatly slower than controls.

Also, female participants were expected to be faster in colour-naming and colour-word naming than males. The response times for male and female participants in word reading would not be different.

Methods

Participants

86 undergraduate students (28 male, 58 female) who were enrolled in PY2101 Brain and Behaviour at James Cook University Singapore participated in the experiment as part of the requirement to complete the module.

Materials

The test consisted of three stimulus sheets. Each stimulus sheet consisted of a list of 100 items, presented in five columns of 20 items, and printed on white A4 paper. Items on the Word page were the words “red”, “green” and “blue”, arranged in random sequence and printed in black ink. No word appeared consecutively in a column. Items on the Colour page were sets of four Xs, printed in red, green or blue ink, and in random sequence. No colour appeared consecutively in a column. Items on the Colour-Word page was created with words from the Word page, and printed in colours from the Colour page. None of the words matched the ink colours they are printed in. The experimenters’ own mobile phones were used to take time.

Procedure

This is a 3 (W, C, CW) X 2 (Female, Male) mixed-design experiment. The experimenters and participants were fully aware of the experimental conditions. The students grouped themselves into pairs to be experimenter and participant.

Each pair was issued with a set of stimuli for the three trials. All participants completed the experiment in the same order – Word, Colour, and Colour-Word. For the Word trial, the participants were required to read the words aloud. For Colour and Colour-Word trials, they had to name the ink colours aloud.

The participants were instructed to go down the list as quickly as possible, and correct the errors in their responses before continuing. The experimenters would point out any error that was overlooked. There was no time limit for the experiment.

The experimenters recorded the time. Timing started with the experimenter saying “BEGIN”, and stopped immediately after the participants had finished all columns, regardless of whether they said “STOP”. After the participants had gone through all three trials, they swapped roles with the experimenters, and repeated the whole procedure.

Data Analysis

Three response times were yielded for each participant. The mean response times and standard deviations were calculated for each trial, and by gender. Stroop Interference is computed as the difference between the time needed to complete CW and the average time needed to complete W and C.

Results

Table 1 summarises the results from the three trials. The mean response times in the three conditions for all participants were compared to examine the Stroop interference. For the control trials, word-reading (M = 43.5, SD = 8.5) resulted in faster response times than colour-naming (M = 58.7, SD = 11.2). For the test trial, participants responded much slower in naming the ink colours of incongruent words (M = 99.1, SD = 20.9) than in colour-naming control.

The mean response times in the three conditions for male and female participants were compared to examine gender differences in the Stroop interference (Table 1). There was no apparent gender difference in word-reading or colour-naming in the control. Male participants (M = 103.9, SD = 18.3) took longer to respond in the colour-word task than female (M = 96.8, SD = 21.8). The interference observed among male participants (53.0 s) was greater than female (45.7 s).

Discussion

Overall, the results seemed to be consistent with the hypotheses. The results showed reading words was faster than naming colors. This finding is supported by Cattell’s (1986, cited in MacLeod, 1991) work. Word-reading is more practised and hence more automatic and faster than color naming.

The Stroop effect was also observed. Participants responded much slower in the test trial than in the colour control trial. Automatic word-reading precedes and conflicts with color-naming, hence leads to slower and erroneous responses to the colour-word stimulus.

Contrary to expectations, there was no strong evidence to support gender difference in word-reading or colour-naming. This result is inconsistent with robust findings that females have superior color-naming skill (Golden, 1974; Jensen, 1965; Ligon, 1932; and Stroop, 1935, all cited in Alansari & Baroun, 2004).

Also, Alansari and Baroun (2004) presented an interesting argument that females’ shorter time latencies on Stroop tasks were due to their faster color-naming ability, rather than to better performance in overcoming the Stroop interference. The findings from this study suggest otherwise. Despite being equal or slightly slower in colour-naming in the control trials, females outperformed males in the colour-word test.

On the other hand, as expected, male participants took longer to respond in the colour-word task than female, therefore greater interference. The large difference observed between naming colours of non-words (C) to words (CW) supports female superiority in the Stroop test. These results concur with prior findings that females show less cognitive interference than males did (Dash & Dash, 1987; Golden, 1974; Pati & Dash, 1990; Peretti, 1971; and Sarmany, 1977, all cited in Alansari & Baroun, 2004).

Areas of Improvement

There are a number of areas of the experiment that can be improved. At the basic level of experimental procedure, the sequence of the trials was not counterbalanced. In addition, the second participant could be confounded as familiarity or when fatigue effect sets in, as he or she had played the role of the experimenter in the first round of trials.

Besides the lack of means testing, the sample size should be larger. According to Roivainen (2010), sample sizes should be more than 100 participants for valid interpretation of results.

The number of errors made by the participants during the trials should be recorded, as accuracy might be traded for speed. Most participants spontaneously corrected themselves when they noticed an error, which indirectly corrected scores for poor accuracy. Despite so, a reading with unusually high number of errors would raise validity issues. The inclusion of practice items and exclusion criteria for data analysis would help to eliminate these outliers.

When reading or color-naming, some participants tended to use their hands to track their progress down the lists. This seemingly simple act may confound with the true effects of Stroop interference. Prior research has revealed spatial processing is enhanced near the hands, potentially benefiting several processes involved in reading (Abrams et al., 2008; Reed et al., 2006; Schendel & Robertson, 2004, all cited in Davoli, Du, Montana, Garverick & Abrams, 2010).

In conclusion, the findings of this study are insufficient to substantiate gender difference in Stroop interference. Despite so, the results corroborate the findings of other studies carried out to examine performance among genders on Stroop tests. It is necessary to conduct further research that takes into account the above mentioned limitations.