, PSYC2092, coursework PDF

Preview

Summary

coursework essay...

Description

1

Does the numerical Stroop task affect the automatic processing of irrelevant stimulus? P2447043 PSYC2092 De Montfort University

2 Abstract The following experiment investigated the automatic processing of irrelevant stimulus, with the congruity effect and the numerical Stroop task. The experiment was performed with the use of the software PsychoPy (Peirce et al.,2019), and the study included one participant. The findings were in line with previous literature and concluded that reaction times in congruent trials are faster than incongruent trials. Furthermore, the results of the neutral condition also support the previous stated fact, as the reaction time was slower than the congruent trials, but still faster than the incongruent trials.

3

Introduction Can we see a difference between the visual size of a number and actual numerical values? Automatic processing of irrelevant stimulus is a well-researched topic, in particular the congruity effect seems to be an outcome of the automatic processing of irrelevant stimulus. The numerical Stroop effect investigates this idea and demonstrates the possible relationship between numerical values and physical sizes (Liu, Wang, Corbly, Zhang & Joseph, 2006). With the Stroop task, participants are asked to make comparative judgements about pairs of numbers presented to them (Heine et al.,2010). A finding that is often replicated across literature is the fact that in Stroop contexts, the congruence of relevant and irrelevant stimulus, affects the participants performance in tasks (Heine et al.,2010). Two possible explanations are the perceptual-attention hypothesis and the response-suppression hypothesis (Kadosh et al.,2011). The perceptual-attention hypothesis is based on the fact that the distraction caused by irrelevant information is attenuated when it is repeated (Kadosh et al.,2011). The response-suppression hypothesis is based on the fact that the reduction of the congruity effect after irrelevant information repetition is due to sustained suppression of the response that was activated from the irrelevant information (Kadosh et al.,2011). This information is derived from the activation-suppression hypothesis. The activation-suppression hypothesis describes suppression of irrelevant information after its initial activation, and this mechanism can explain the decrease of the congruity effect with increasing response times (Kadosh et al,2011).

Heine et al., (2010) investigated the automatic processing of irrelevant stimulus among children. The aim was to provide further insights into the development of automatized numerical processing in children. Additionally, the study explored the size congruity and numerical distance effects under Stroop conditions in three groups of primary school children with different math achievement skills (Heine et al.,2010). The results of their study showed that the congruity effect appeared in the children, subject to their version of the Stroop task, in all three groups of high achievers, low achievers and normal achievers in math (Heine et al.,2010).

Kadosh et al., (2011) investigated the automatic processing of irrelevant stimulus in students, with a focus on the two alternative hypotheses mentioned above. The aim of this particular study was to assess these two hypotheses, using a numerical Stroop task (Kadosh et al.,2011). The study consisted of 21 students and the results suggested that the congruity effect is significantly reduced when the response sequence of the irrelevant dimension is repeated (Kadosh et al.,2011). The results showed that the responsesuppression hypothesis is the most relevant explanation to the congruity effect, and with the use of it, the

4 congruity effect can be avoided all together (Kadosh et al.,2011). The results also shed light on the level of interaction between numerical magnitudes and physical size as well as the effect of variability of responses and stimuli on automatic processing (Kadosh et al.,2011).

Liu et al., (2006) investigated the role of the parietal cortex and the congruity effect. The aim of the study was to get a closer look at the relationship between the parietal cortex and cognitive functioning’s such as the numerical Stroop effect (Liu et al.,2006). fMRI was used as the method, as well as a version of the numerical Stroop task to see whether the parietal cortex was activated (Liu et al.,2006). The participants consisted of 12 native English speakers and the results concluded that participants responded more slowly to incongruent pairs, and the inferior parietal cortex showed common and distinct patterns of activation for both attentional selection and number comparison (Liu et al.,2006). The activity was modulated by a Stroop-like interference effect and the distance effect (Liu et al.,2006).

With previous literature considered, the following experiment aimed to look at the automatic processing of irrelevant stimulus, using the numerical Stroop task. The experiment considered the following hypotheses: Hypothesis 1: Response time in incongruent trials will be significantly slower than in congruent trials. Hypothesis 2: Response time in congruent trials will be significantly faster than in incongruent trials. Hypothesis 3: Response time in neutral trials will not differ much across neutral trials, in comparison to the incongruent and congruent trials.

Methods Participants: A 22-year-old female undergraduate student was used as participant. Design The experiment manipulated the three different conditions, congruent, incongruent and neutral. The dependent variable was the reaction times in each condition for the participant. For this experiment, only the accurate response times for each trial was measured. The experiment followed a between-participants design and the participant was randomly assigned to each trial.

Materials The experiment was presented using PsychoPy 3.2.3 software (Peirce et al.,2019), on a HP computer. Trials started with the presentation of a fixation cross for 1 second and a blank screen

5 for 0.2 seconds. The Participant was presented with numbers of different sizes, numerically and visually, each target was present until the participant pressed either the right or left arrow key. The different trials for each condition was presented in PsychoPy with the use of an excel file with the main trial files, and one excel file with the practice trials. Each excel file was provided with a column containing information for each condition, the name of the column to find the correct answers from was resp.corr, and the column with the information regarding the dependent variable was called resp.rt. The PsychoPy script contained five routines in total: the feedback routine, the instruction routine, the practice routine, the trial routine and the final thanks routine. The instruction routine, the practice routine and the trial routine consisted of keyboard components. Each keyboard component was named differently, with the trial routine being named resp, and the instruction named ready and the practice routine named ready1. For the instruction routine, participants were instructed to press any key to start the experiment. For the practice routine and the trial routine, the use of the left and right arrow keys was instructed to the participants, to register the responses. Correct responses received the feedback “Correct” as well as the response time, and incorrect responses received the feedback “Oops that was wrong”. The feedback was present for one second, before moving on to the next trial. The main trials were not given any feedback. Participants completed 14 practice trials. There were four practice trials for each level of the condition, and in half the practice trials the target was present in the set. Numbers presented as well as the target position in the set varied across practice trials. Participants were allowed to answer at the same time as the target presentation (0.75 seconds into the trial) and the next trial began immediately following a response. The inter-stimulus interval for this experiment was 500ms.

Procedure The experiment was performed using the software PsychoPy (Peirce et al.,2019), on a HP computer. The participant was instructed to download the trial and practice files, along with the Psychopy script, and run the experiment. The participant was then asked to put in their unique student number before running the experiment, in order for the data to be recorded. The participants task was to indicate which number is bigger or smaller, as each trial represented numbers that were of different sizes, visually and numerically. The experiment commenced with 14 practice trials, where participants were provided with feedback. The experiment consisted of

6 three conditions, congruent, incongruent and neutral. In a congruent trial, participants were presented with the right number in the right size. In an incongruent trial, participants were presented with a smaller number in a big size and the bigger number in a smaller size. In a neutral condition, participants were presented with two numbers of visually equal size, but of different numerical size. There were 14 practice trials and 97 main trials in total, with trials separate for each condition (i.e. 33 trials for the congruent condition, 31 for the incongruent condition, 31 for the neutral condition). Participants were instructed to use the left and right arrow keys to submit their answer. The trials were presented randomly, and each trial was repeated once during the practice trials and one time during the main trials. Following the completion of the experiment, the raw data file was accessed through an excel file created by the software. The practice trials were removed and the cells were sorted into correct answers, labelled resp.corr in the excel file. Following that, all the incorrect responses were removed, and the data was sorted into each condition, and the averages and standard deviations were calculated, refer to appendices B, C, D, E, F and G.

Data collation For this experiment, the practice trials were removed, and the cells were sorted by the correct answers, labelled as resp.corr in the excel file. After sorting according to correct answers, the incorrect answers were removed. Finally, the cells were sorted by condition.

Results The results from the experiment was analysed using different tools in Microsoft excel, see Appendix A. To calculate the average and the standard deviation, the Subtotal tool was used to calculate averages and standard deviations for each condition, as seen in appendices B, C, D, E, F and G. The independent variable was the condition, which consisted of three levels: congruent (M=0.83, SD=0.34), incongruent (M=1.06,SD=0.36) and neutral (M=1.01,SD=0.34). The dependent variable was the reaction times for each condition, displayed in the means and standard deviations. The results support previous literature, that reaction times in congruent trials are faster than incongruent trials. Furthermore, the results of the neutral condition also support the previous stated fact, as the reaction time was slower than the congruent trials, but still faster than the incongruent trials.

7

Discussion The experiment examined the automatic processing of irrelevant stimulus using a numerical Stroop task and consisted of one independent variable with three levels and one dependent variable. The independent variable was the condition and the three levels consisted of congruent, incongruent and neutral trials, and the dependent variable was the reaction times for each condition. The results indicated that reaction times in congruent trials were faster than incongruent trials. The neutral condition confirmed this idea, as the reaction time in the neutral condition was slower than the congruent trials, but faster than the incongruent trials. The results supported the first and second hypotheses, however, hypothesis three was not supported.

The findings from this experiment, are in line with previous research. In particular, the results are in line with Heine et al., (2010), where the congruence of relevant and irrelevant stimulus affects the participants performance in tasks (Heine et al.,2010). The study conducted by Heine et al., (2010) had some strengths, but also some weaknesses. The research made into this study suggested that a focus on the interaction of both effects, is a promising approach instead of solely looking for size congruity effects, which is the reason the Stroop task was implemented (Heine et al.,2010). The fact that the methodology is the same as the current experiment, makes Heine et al’s (2010) study promising. Furthermore, the Stroop task implemented in this study allowed for a separate, in-depth analysis of the respective influences of congruity and numerical distance in children’s Stroop performance (Heine et al.,2010). The weakness considers the fact that the main focus is children. On the one hand, the results of the study have been implemented across previous literature and this could perhaps suggest that the effect is similar across children and adults, making it relevant for the use for the current experiment (Heine et al.,2010). On the other hand, it is hard to use that as a sweeping statement, without having research that compares children and adults. However, the current experiment only included one participant, which makes the results questionable as one participant cannot represent a whole population. Thus, previous research is important for the implication of the results. Although Heine et al., (2010) study only included children, it still represents more participants than the current experiment.

8 As mentioned previously, there are other explanations to the results of this experiment, which could perhaps indicate the overall implication of the results. Kadosh et al., (2011) assessed the perceptual-attention hypothesis and the response-suppression hypothesis. The study is relevant to the current experiment, because the results are the same, the response times are slower in incongruent trials than congruent trials. However, with the use of the hypotheses, this study shows how to avoid the congruity effect, which makes it interesting as it provides an explanation (Kadosh et al.,2011). The particular strengths to this study were the fact that the participants were relevant to the current experiment (Kadosh et al.,2011). In comparison to Heine et al.,(2010) study where the focus was on children, this study presents information that is relevant to the group of participants in the current experiment. Additionally, the assessment of the two hypotheses, is another strength because it is important as a possible explanation to the congruity effect and it shows cases where the congruity effect can be avoided. The weaknesses of this study are part of its strengths, the fact that it is an assessment of the two hypotheses, rather than looking at the congruity effect across participants, makes it less relevant to the current experiment. However, this could be considered as a minor weakness, as it provides an actual implication of the result of the current experiment.

Furthermore, Heine et al., (2010) provided an alternative explanation to the phenomena, meaning that the posterior parietal cortex has been identified as possibly being related to a number of different cognitive functions. The cognitive functions that are said to be related to the posterior parietal cortex, have a few aspects in common. For instance, the fact that they draw on magnitude representations in more general terms (Heine et al.,2010). This is a cross-domain representational overlap that might be one of the reasons for the interaction of numerical and physical size information that can be observed in numerical Stroop tasks (Heine et al.,2010). The study conducted by Liu et al., (2006) explored the matter further and the strengths of the study was the explanation of a possible neural correlation with the congruity effect. The study shows evidence for this as the parietal cortex was active during the Stroop tasks, as it appeared on the fMRI scans of the participants (Liu et al.,2006). However, apart from the cross-domain representation, three studies demonstrate that the resolution of conflict between the two dimensions of numerical and physical size might not just be limited to early processing stages (Heine et al.,2010). Furthermore, as the current study did not have a neurological component,

9 this study is not directly relevant, but it is an interesting approach. Nevertheless, as the current study is limited to one participant’s data, it does not exclude the role of the parietal cortex and the congruity effect, as it could have been implemented in a study with a larger scale of participants.

In conclusion, the current experiment shows strengths and weaknesses. The findings replicate previous literature, but there is a need of further data in order to understand the overall implications of the results.

References Cohen Kadosh, R., Gevers, W., & Notebaert, W. (2011). Sequential analysis of the numerical Stroop effect reveals response suppression. Journal Of Experimental Psychology: Learning, Memory, And Cognition, 37(5), 1243-1249. doi: 10.1037/a0023550 Heine, A., Tamm, S., De Smedt, B., Schneider, M., Thaler, V., & Torbeyns, J. et al. (2010). The Numerical Stroop Effect in Primary School Children: A Comparison of Low, Normal, and High Achievers. Child Neuropsychology, 16(5), 461-477. doi: 10.1080/09297041003689780 Liu, X., Wang, H., Corbly, C., Zhang, J., & Joseph, J. (2006). The Involvement of the Inferior Parietal Cortex in the Numerical Stroop Effect and the Distance Effect in a Two-digit Number Comparison Task. Journal Of Cognitive Neuroscience, 18(9), 1518-1530. doi: 10.1162/jocn.2006.18.9.1518 Peirce, J., Gray, J., Simpson, S., MacAskill, M., Höchenberger, R., & Sogo, H. et al. (2019). PsychoPy2: Experiments in behavior made easy. Behavior Research Methods, 51(1), 195203. doi: 10.3758/s13428-018-01193-y

10

Appendices Appendix A Raw data file

Appendix B Averages for the congruent condition

11

Appendix C Averages for the incongruent condition

Appendix D

12 Averages for the neutral condition

Appendix E Standard deviation for the congruent condition

13 Appendix F Standard deviation for the incongruent condition

Appendix G Standard deviation for the neutral condition

14

15...