Many disorders show inhibition deficits such as schizophrenia, ADHD, Tourette’s syndrome, and obsessive-compulsive disorder. Inhibition deficits include responding before the task is understood, losing attention and failing to correct inappropriate responses. Patients with Schizophrenia show good performance on stop signal tasks, but there is reduced activity in the mesial prefrontal brain regions during inhibition (hypofrontality) (Rubia et al 2001).
Children with ADHD respond quickly and inaccurately on the stop signal task compared to normal controls. Schachar & Logan (1990) found this reflected impulsivity and lack of inhibition, especially if there was a long stop signal delay. This has been correlated with high ratings of hyperactivity in the classroom (Pliszka et al 1997) and improvements in behaviour, and academic performance in ADHD children being treated with methylphenidate strongly associates with improvements in inhibitory control (Tannock et al 1989).
Many factors affect our ability to inhibit responses e.g. timing, stimulus type, and motivation. If inhibition has an effect, the longer the stop signal delay, the harder it is to inhibit a response. If there is a short stop signal delay, one can change their response as the stop signal is presented before a decision is executed. Also performance increases if positive rewards are associated. No rewards lead to worse inhibitory response especially in Schizophrenics and ADHD patients. Konrad et al (2000) found that rewards on the stop signal task, brought children with ADHD up to the performance level of normal controls. Different types of stimuli may also affect inhibitory responses as it is possible that different parts of the brain may be used for different stimuli e.g. image vs. letters. If the difficultly of the task increases the time to make a response decision, the subject has more time to inhibit their response (Logan & Cowan 1984). Also some tasks have been found to be harder to inhibit than others. Logan (1982) tested skilled typists on the stop signal task and found that it was harder to inhibit the space after the word ‘the’ and put familiar endings on verbs e.g. ‘ed’ and ‘es’, i.e. the most familiar actions.
This experiment is looking at the affects of varying the stop signal delay to see what the temporal characteristics of control are. The effect of easy stimulus (single letters) and difficult stimulus (five letter words) on inhibition to see if difficulty affects decision time is also analysed. The experiment follows a similar procedure to Logan & Cowen’s (1984) stop signal task, however this experiment uses visual stimuli. The experimental hypotheses are:
- Increasing the stop signal delay will decrease the ability to withhold responses.
- Increasing the stimulus difficulty will increase the ability to withhold responses.
- Increasing stimulus difficulty and increasing the stop signal delay will together affect the ability to withhold responses.
Method
Pilot Experiment
A pilot experiment revealed ceiling effects with accuracy being too high at all the stop signal delays (200ms, 250ms, 300ms, and 350ms) as the stop stimuli were presented before subjects could even decide or react. This was modified and re-tested (results in the appendices) before collecting the full data to 250ms, 350ms, 450ms, 550ms (based on timing tests, listed in appendices) making it less predictable for subjects.
Subjects
Twenty participants took part. 17 were female and 3 were male. All of them were first year undergraduates at the University of Nottingham, England, who volunteered to participate and were selected randomly. The mean age of the participants was 19 years however the age range was from 18 years to 21 years.
Design
The experiment used a within subjects design as each participant was assigned to all the conditions. All participants completed the same task, but the order in which each block of stimuli presented was randomised by the computer to prevent order effects.
The independent variables were the stop signal delay and stimuli difficulty. The delay varied between 250ms, 350ms, 450ms and 550ms based on the means calculated in the timing test (see appendices). Stimuli difficulty varied between the easy condition (single letters) and the difficult condition (five letter words).
The dependant variable was the percentage accuracy of performance in inhibiting the response to the changes in the independent variable, recorded by the computer.
The experimental hypotheses were:
- Increasing the stop signal delay will decrease the ability to withhold responses.
- Increasing the stimulus difficulty will increase the ability to withhold responses.
- Increasing stimulus difficulty and increasing the stop signal delay will together affect the ability to withhold responses.
Materials & Apparatus
The stimuli in the easy condition were the single letters ‘A’ and ‘B’. The stop signal was the letter ‘S’. The stimuli in the difficult condition were the five letters words ‘bored’ and ‘plain’. The stop signals were the corresponding homophones ‘board’ and ‘plane’. All the words or letters were white on a black background; of Arial font size 30 (see appendices).
Subjects used a Samsung Sync Master 753DFX Personal Computer and keyboard to run the experiment. The experiment was created and performed on E-Studio Programme.
Procedure
The participants took part individually. The participants in their own time watched the computer screen and ran the experiment through E-Studio programme. The participants were shown standardised instructions on the screen of how to carry out the task for 1,000,000ms (see appendices) and asked if they understood what was being asked of them. Then the participants were shown standardised instructions relating to the block they would shortly be presented with either easy stimuli or difficult stimuli. The participants were then presented with a fixation point in the centre of the screen for 2000ms. This was followed by the presentation of a stimulus in the centre of the screen (randomly selected, easy or difficult) which remained until the participant responded by pressing a corresponding key on the keyboard (‘1’ for ‘A’ or ‘bored’ and ‘2’ for ‘B’ or ‘plain’). After the key was pressed the stimulus disappeared and was replaced by another stimulus. On some trials (randomly selected), the stimulus was followed by the corresponding stop signal. This would be presented 250ms, 350ms, 450ms or 550ms after the stimulus (randomly varied between blocks). When the stop signal was presented the participants had to refrain from pressing any key i.e. inhibit their response. The participant had to complete 8 blocks (4 easy, 4 difficult), with 20 trials on each block, in a random order. The computer recorded the participants’ percentage accuracy in responding correctly.
Results
The means (and standard deviations) of the data are shown in the table below. A graph with standard error bars and the raw data is listed in the appendices.
Table 1: Table of average mean response accuracy (and standard deviations) across all conditions
The means show that there is a decrease in accuracy on both easy and difficult stimuli from 250ms (easy stimuli mean=0.82, difficult stimuli mean=0.88) to 550ms (easy stimuli mean=0.50, difficult stimuli mean=0.55) and that the accuracy was generally lower on the easy stimuli. To see if these differences are significant, inferential statistics need to be done.
We tested the assumptions of ANOVA (analysis of variance), i.e. the homogeneity of variance and normality. These were violated; therefore we conducted a reciprocal transformation on the data. However, the assumptions were still violated therefore we are relying on the robustness of ANOVA of our original data. This may have implications when interpreting the results. Sphericity was also violated therefore we used the Greenhouse-Geisser correction.
A 2 (easy stimuli and difficult stimuli) x 4 (250ms, 350ms, 450ms, 550ms) repeated measures ANOVA was conducted on participants’ accuracy scores. The main effect of stimulus difficulty on participants accuracy was not significant (F1, 35=1.425, MSe =6.368, p=0.247). The main effect of stop signal delay on participants accuracy was significant (F2, 35=33.243, MSe=3.831, p=0.001). Accuracy was higher for 250ms (mean=0.85) than for 550ms (mean=0.5). Bonferroni corrected paired t-tests (p=0.193) revealed that there was not a statistically significant difference in accuracy scores when the stop signal was presented 250ms as opposed to 350ms after the go stimulus (mean=0.85 at 250ms, mean=0.8 at 350ms). All of the other differences were statistically significant. There was no significant interaction between delay and difficulty on participants accuracy (F2, 35=0.505, MSe=2.556, p=0.594).
Participants were more accurate at inhibiting responses with a shorter stop signal delay. Task difficulty had no effect on inhibition and there was no interactive effect.
Discussion
The means show a decrease in accuracy on both easy and difficult stimuli from 250ms (easy stimuli mean=0.82, difficult stimuli mean=0.88) to 550ms (easy stimuli mean=0.50, difficult stimuli mean=0.55) and accuracy was lower on the easy stimuli. However the inferential statistics found that the difference in stop signal delay was the only significant finding. The experimental hypothesis that increasing the stop signal delay will decrease the ability to withhold responses can be accepted. However the other hypotheses must be rejected: Increasing the stimulus difficulty will decrease the ability to withhold responses, and increasing stimulus difficulty and increasing the stop signal delay will together decrease the ability to withhold responses, as these were not significant.
The significant effect of the stop signal delay can be explained using Logan and Cowan (1984)’s Horse Race Model. The stop signal starts a stopping process which ‘races’ with the thought processes already running in reaction to the go stimulus. If the stopping process wins, inhibition occurs. If another process wins, the action runs on until completion (figure 1). If there is a long stop signal delay (e.g. in this case 550ms), this reduces the probability of inhibiting a response and if there is a short stop signal delay (in this case, 250ms), the response will nearly always be inhibited. By 550ms, in both easy and difficult conditions the thought processes were completed on approximately half the trials (easy stimuli mean=0.50, difficult stimuli mean=0.55), therefore making it harder to inhibit the response than at 250ms (easy stimuli mean=0.82, difficult stimuli mean=0.88).
The effect of task difficulty could have been explained by different types of stimuli use different parts of the brain. If the difficultly of the task affects decision time, then the subject has more time to inhibit their response (Logan & Cowan 1984) e.g. at 250ms the response to an easy stimulus is decided, but not for a difficult stimuli therefore it is easier to inhibit. By 550ms response is decided to both stimulus therefore it is now equally hard to inhibit. However even though the mean accuracy scores showed that accuracy was lower on the easy stimuli, inferential statistics showed this was not significant, suggesting that letters and words use the same parts of the brain. Konishi et al (1999) found localisation of activity on the go/no-go task and WCST in the posterior right inferior frontal sulcus, suggesting this is the common central mechanism for inhibition, as it prevents both responses, which would support these findings. However Gavaran et al (1999) found inhibition is more distributed throughout the brain, although these areas are primarily right ventral frontal lobe regions.
The lack of a significant finding could be due to experimental design flaws. The stimuli may could be too similar or the difficult condition not significantly more difficult than the easy condition. Perhaps words use the same part of the brain as letters; therefore we are not actually testing a different part of the brain. Further research should account for this by changing the stimuli e.g. increasing the number of stimuli or using words and pictures.
Another limitation is individual differences in strategy. As participants go through the task they refine their strategy, i.e. they withhold their responses to see if a stop signal is presented before deciding, which would make the results less reliable as it would not be an accurate reflection of natural inhibition. Another strategy would be not to answer at all giving 100% accuracy on all of the stop signal trials. One way of avoiding the use of strategies is to use a tracking algorithm. This would change the stop signal delay as the experiment progressed, by monitoring participant’s responses and changing the delay accordingly e.g. if the participant slows down when realising there is a stop signal, the algorithm would increase the delay.
The results should be interpreted carefully due to the violations of the assumptions of ANOVA; therefore it is possible that the effects found are not correct. Perhaps in a replication of this study, the number of participants should be increased to reduce outliers which could possibly produce an effect that is more reliable and significant.
Further expansion of this study could be to look at the effects of negative and positive motivation on inhibitory control. For example, Konrad et al (2000) found that rewards on the stop signal task, brought children with ADHD up to the performance level of normal controls. Or looking at the inhibition of control in motor skills e.g. avoiding a falling object when walking a straight line, rather than cognitive.
Although this experiment did not provide any significant result for task difficulty, plenty of studies have been done to provide evidence for the factors affecting inhibition. Understanding inhibition would have great implications not only for the psychological field of automatic behaviours but for the general public. Understanding how or what makes us inhibit our responses could help understand and treat many disorders e.g. ADHD, Schizophrenia and frontal lobe dysfunction therefore continuing research into other factors to help explain the relationship is important and relevant.
References
Logan, G., Cowen, W., and Davis, K. (1984). On the Ability to Inhibit Simple and Choice Reaction Time Responses: A Model and a Method. Journal of Experimental Psychology: Human Perception and Performance Vol. 10, 276-291
→ Lappan & Eriksen (1966), Lisberger et al. (1975), Logan (1981) obtained from the above source.
Logan, G., and Cowen, W. (1984). On the Ability to Inhibit Thought and Action: A Theory of an Act of Control. Psychological Review 91, 295-327
→ Craik (1947 1948), Slater-Hammel (1960), Henry & Harrison (1961), Shiffrin and Schneider (1977), Lisberger et al (1975), Logan (1982) obtained from the above source.
Schachar, R. & Logan, G. (1990) Impulsivity and Inhibitory Control in Normal Development and Childhood Psychopathology. Development Psychology 26, 710-720
→ Case & Globerson (1974), Tannock et al (1989) obtained from the above source.
Konishi, S., Nakajim, K., Uchida, I., Kiyo, H., Kameyama, M., & Miyashita, Y. (1999) Common Inhibitory Mechanism in Human Inferior Prefrontal Cortex Revealed by Event-Related Functional MRI. Brain 122, 981-991
→ Iversen & Mishkin (1970) obtained from the above source.
Garavan, H., Ross, T. & Stein, E. (1999) Right Hemisphere Dominance of Inhibitory Control: An Event-Related Functional MRI Study. Proceeding of the National Academy of Science USA 96 8301-8306
Rubia, K., Russell, T., Bullmore, E., Soni, W., Brammer, M., Simmons, A., Taylor, E., Andrew, C., Giampietro, V. & Sharma, T. (2000) An fMRI Study of Reduced Left Prefrontal Activities in Schizophrenia during Normal Inhibitory Function. Schizophrenia Research 52 47-55
Pliszka, SR., Liotti, M. & Woldorff, MG. (2000). Inhibitory Control in Children with Attention-Deficit/Hyperactivity Disorder: Event-Related Potentials Identify the Processing Component and Timing of an Impaired Right-Frontal Response-Inhibition Mechanism. Biological Psychiatry 48 238-246
→ Sergeant and van der Meere (1989), Pliszka et al (1997) obtained from the above source.
Konrad K, Gauggel S, Manz A, & Scholl M. (2000). Lack of inhibition: a motivational deficit in children with attention deficit/hyperactivity disorder and children with traumatic brain injury. Neuropsychology Development Cognitive Section C Child Neuropsychology 6 (4) 286-96.
Appendices
- Results from the Pilot Test
- The stimuli and ‘stop’ signals used in the experiment
- The instructions presented to the participants.
- Graph of results
- Raw Data
- Averages, Standard Deviation and Error
- SPSS output