This page requires JavaScript be enabled in your browser to work properly!

interactive language learning
British Academic Written English (Life Sciences) your name:

"Distraction Data"-Data Analysis Project


Previous research has shown that boys of primary school age are involved in a greater number of road traffic accidents than girls. It is possible that the reason for this is due to a difference in the development of attention skills between boys and girls of this age range. In this experiment we want to investigate differences in the attention skills between boys and girls between the ages of four years and nine years, and observe how these attention skills are affected by a background distraction. We predict that boys will perform worse than the girls at an attention task when being distracted by a background event. As attention skills are known increase with age, it is also predicted that performance by both girls and boys in both the distraction and no-distraction conditions will increase with age.



The participants in this experiment were 96 children from a local primary school. The children were divided into three groups based on age, 4-5 years, 6-7 years and 8-9 years. There were equal numbers of male and female participants in each group.


The stimulus used in the experiment was the 'Opposite Worlds' attention test, a sub-test in the "Test of Everyday Attention for Children" (TEA-ch). The TEA-ch was developed by Manly, Anderson, Nimmo-Smith, Turner, Watson and Robertson (2001) for the purpose of measuring various types of attention skills in children. The test was specifically developed to be a pure way of measuring attention, and this was achieved by ensuring that the tasks within it placed minimal load on cognitive functions other than attention, as well as using structural equation modelling to control the variables being measured. As such, it can be said to have a satisfactory level of validity (Manly et al., 2001). In addition, test scores were analysed in development to ensure a high level of validity, and reliability has been established using test-retest procedures (Manly et al., 2001). Thus, we can be confident that the TEA-ch is both a valid and reliable test of attention. The 'Opposite Worlds' sub-test that we are using was designed to measure attention at a verbal inhibition task.


Children of all age groups were first required to complete the baseline test, as is normal for the test. In the baseline test, children are asked to read out a list of one's and two's in a path. Children were then required to complete the 'Opposite Worlds' test a first time with no distraction. In this part of the test, children must say "two" when they see a one and vice versa, and hence are required to inhibit the natural verbal response to the presented digit.

All children were then required to complete the test a second time, and were told that on completion of the test they would be allowed to watch an interesting cartoon on the television. However, during this second run of the test, the television was switched on remotely halfway through the test. This provided the background distraction event. The scores for the two conditions (distraction and no distraction) were calculated by subtracting the average time taken to complete the baseline task from the average time taken to complete the task in each condition. A higher score denotes a poorer performance.


The independent variables for this test were the age and sex of the children taking part. The three levels for age were as described in the participants section. As sex and age were the independent variable, this actually deems this experiment a quasi-experiment. The dependent variable was the amount of time taken to complete the tasks in the distraction condition and the no-distraction condition.


We wanted to know whether the age or sex of primary school children had any affect on their test-scores for the 'Opposite Worlds' subtest in the test of everyday attention for children (TEA-ch). Prior to analysis, the data was screened for outliers, missing data and coherence with the basic assumptions of the statistical test used. Two outliers were identified, one in 6-7 boys group of the no-distraction condition and the other in the 8-9 girls group of the distraction condition. As there is no reason to doubt the validity of these results, they will be retained for the analysis. However, the residuals from the ANOVA on the untransformed dataset were found to be non-normally distributed, and so the data was square-rooted for analysis. No other breaches of test assumptions were found.

Our first prediction was that boys would perform worse than girls on the 'Opposite Worlds' task when being distracted by a background event, and hence our null hypothesis was that sex would have no effect on performance in the distraction condition. Table one suggests that boys took longer than girls to complete the task in the distraction condition in all age groups. A two-way ANOVA carried out on the independent variables age(3) and sex(2) confirmed that there was a significant main effect of sex on performance F(1,90)=36.16, p<.01. This supports our hypothesis. Table two suggests that boys were affected to a greater negative extent than girls by the distraction, and a two-way ANOVA performed on the difference scores confirmed this F(1,90)=4.56, p<.05. In addition, a main effect for sex in the no distraction condition was also found by a two-way ANOVA, F(1,90)=11.82, p<.05. Table suggests that this interaction could mean that boys' attention skills develop at a slower rate than girls.

Our second hypothesis was that, as it is known that attention skills increase with age, so performance at the Opposite World task should improve with age in both conditions. Hence our null hypothesis was that there would be no main effect of age. Table one suggests that there was a steady decrease in time taken, and hence increase in performance, as the children increased in age in the distraction condition, and table three suggests the same trend in the no-distraction condition. A two-way ANOVA was carried out for both conditions on the independent variables age(3) and sex(2). The results confirmed that there was a significant main effect of age on the performance at the tasks in both the distraction condition F(2,90)=9.53 p<.01 and the no-distraction condition F(2,90)=16.04, p<.01. These findings support our second hypothesis.

The two-way ANOVA's conducted above also found a significant interaction between age and sex in the distraction condition F(2,90)=3.62, p<.05, whilst finding that there was no significant interaction in the no-distraction condition F(2,90)=.13, p<.10. Table one suggests that this interaction means that boys do not improve at the 'Opposite Worlds' task between the two higher age ranges, i.e. 6-7 years and 8-9 years.

Residuals from both ANOVA's were screened for normal distribution and systematic variation with other variables. Correlation with original baseline scores was found to be significant for residuals from both distraction r(95)=0.21, p<.05 and no-distraction r(95)=.46, p<.01 conditions. This is expected to a certain extent, and so the ANOVA findings were maintained.


We predicted that, as boys are observed to be involved in a greater number of traffic accidents than girls, and as this finding was probably due to poorer attention skills of boys than girls, boys should perform worse at an attention task. Our results support this prediction, showing that boys do indeed perform more worse girls at the attentions tasks, both in the distraction and no-distraction conditions. Our findings also revealed that a distraction had a greater negative impact on the boys than the girls, something that should be taken into account when considering road safety. For example, whilst a distracting billboard and a conversation with a friend might impede girl's abilities to concentrate on the safety of the road, the same distractions would cause a more serious problem for boys. As a result, and coupled with the fact that our findings suggest that boys already have poorer attention skills than girls, boys are even more likely to make a mistake in judging the safest time to cross the road under distracting conditions.

Our second hypothesis was that performance on the attention tasks would increase with age, as attention skills are known to improve with age. Our findings showed this trend, however they also showed something that was not expected, that there was an interaction between age and sex in the distraction condition. The direction of the results suggested that boys between the ages of 6-7 years and 8-9 years did not progress in the development of their attention skills. We could interpret this as meaning that, not only are boys already behind girls in the attention skills at the age of four, they may also be slower at developing these skills. However, the girls 8-9 group on this condition was found to contain an outlier, and this should be considered before accepting this finding, as it would have increased the mean girls score for this group. Nevertheless, an interesting experiment to follow on from this would be to see if boys eventually accelerate in the development of their attention skills to match those of girls. Assuming that adult males and females have equal attention skills, we would expect this acceleration to occur somewhere during later childhood. Such a finding might be important for general education of the different sexes.

The other finding from our analysis was that the residuals from both conditions correlated with the original baseline task score. We can interpret this as meaning that the baseline task score is a useful predictor of the score obtained in the actual 'Opposite Worlds' tasks. However, this is not unexpected, as it would follow that someone who performed poorly in the baseline task would perform proportionally poorly in the actual 'Opposite Worlds' task.

We now consider the methodology of the experiment. Firstly, only one sub-test was used from the test of everyday attention for children (TEA-ch), the 'Opposite Worlds' task. As noted earlier, the TEA-ch was designed with different tasks to test different types of attention. In using only the one task, we have essentially tested only one type of attention, which in this case is attention at a verbal inhibition task. Whilst the idea of inhibition in general is appropriate to the domain of investigating traffic accidents, the "walk don't walk" subtest of the TEA-ch was developed for the more specific type of attention that is important in considering whether it is safe to cross a road . Hence, to a certain extent, using only the 'Opposite Worlds' sub-test has reduced the validity of the experiment, and future investigation into the same area might consider using different sub-tests.

Another potential methodological problem was the fact that in the distraction condition, participants were distracted for only half the time, as the TV showing cartoons was switched on only half-way through the task. Had the distraction been present for the whole of the test under this condition, we might have observed even more impediment of attention skills, reflected in higher scores for this condition. However, this should not affect the significance of the outcome of the experiment, it should merely be noted that the findings from the distraction condition are probably not as high as they should otherwise be.

Finally we discuss the generalisability of the findings. It should first be noted that there is a very different motivation for completing the attention task as there is for crossing a road. Whilst children were told that they would be rewarded for completed the task (with a cartoon), this does not compare with the possible consequences of failing to exercise due attention when crossing the road, which could lead to a serious accident. The immediate sensory information from a road crossing context may lead to greater focus of attention on the task at hand. In the attention task presented here, there was no motivation to complete the task successfully, merely to complete it. In contrast, there was no motivation for not finishing the test correctly, as this would not result is finishing the test any faster. So we could assume that the findings are, at least, an indication of the trend of attention between the sexes, even if in real situations involving crossing a road immediate attention might be higher.

In conclusion, we have seen that boys do perform worse than girls at an attention task, and this could be part of the reason that boys are observed to be involved in a greater number of traffic accidents.


  • Manly, T., Anderson, V., Ian, N., Turner, A., Watson, P., & Robertson, I. H. (2001). The differential assessment of children's attention: the test of everyday attention for children (TEA-ch), normative sample and ADHD performance. Child Psychology and Psychiatry, 42, 8 1065-1081. Cambridge: Cambridge University Press.
Please wait for loading ...
Please wait for loading ...
Please wait for loading ...


Related topics in Wikipedia