Camel and Cactus Test – Term Paper

Analysis of the impacts of gender, age and education on an individual’s cognitive functions using the Camel and Cactus Test

Abstract

This report is a study of the impacts of gender, age and education on an individual’s cognitive functions using the Camel and Cactus Test (CCT). The study uses the convenience-sampling method, to draw a sample of 55 participants. The participants completed a CCT, and the result was analyzed using SPSS, with the data presented in tables and figures. From the study, it was determined that age and education are linked to an individual’s cognitive functions, whereas there is no link between gender and cognitive functions. The results are similar to extant studies on the same.

Introduction

The present report outlines the study about the impacts of gender, age, and education on an individual’s cognitive functions. The study adopted the Camel and Cactus Test, which is part of a larger tool called the Cambridge Semantic Memory (CSM) battery test, to find the link between selected demographic properties and cognition. Specifically, the study involved administration of questionnaire to participants selected through convenience sampling. Notably, the CSM test is a collection of tests, but they are all based on similar stimulus items to find out the extent to which semantic knowledge is input and output using different modalities. It uses 64 items under different subcategories, including CCT, which is a measure of semantic association, picture naming, category fluency, word picture matching (word comprehension), and sorting by category. The CCT, which is the focus of the rest of this study, was developed from the Pyramid and Palm trees Test, and like the others, is useful in observing patients suffering from semantic dementia (SD). According to Bozeat et al. (2000), the patients who show SD symptoms usually have bilateral anterior temporal lobe atrophy, although those with mild conditions usually fail to exhibit any useful deficits when tested using Word-Picture matching or Pyramids and Palm Trees Test. As such, the CCT tests can help in determining patients with mild conditions of SD, thereby increasing the chances of early detection and early treatment.

Participants for the CCT test are expected to select correct responses from four items of the same category, and assessment takes both verbal and non-verbal cues, thereby making it possible for the researcher to detect differential impairments across the two different modalities of output and input. This study is important because, as Grundman et al. (2004) clearly note, while previous studies on AD focused on the cause of deficit, as opposed to the stage at which the disease can be detected. Thus, according to Grundman et al (2004), the early stage of the disease is called Mild Cognitive Impairment (MCI), and studies indicate that not all cases of MCI fully develop into full AD. However, there is interest to determine whether AD has any associations with the   Thus, it is timely to carry out an experiment that relates psychometric properties and the prevalence to AD.      

Methodology and participants

 Results

Based on the methodology described above, the data gathered was analysed using frequency distribution, measures of central tendency and analyses of  variance (ANOVA), and the variables included while measuring the performance of participants on the CCT were age, education and gender. The detailed results are presented below.

Table 1: age of participants

Figure 1: Age distribution of participants

Table 2: gender and education of participants

Figure 2: Gender distribution of anticipants

Figure 3: education level of participants

From figures 1, 2 and 3 and tables 1 and 2 above, it is notable that out of 55 participants, 23 (41.8%) were males whereas 32 (58.2%) were females, and 67.3% of them had attained higher level education. Regarding the age of participants, a majority of them were aged between 21 and 40 years although the mean age was 34.2 years. table 3 further shows the number of participants segmented according to age, gender, and education level in that order, and it is seen, for instance, that six males aged between 21 and 40 years had higher level of education.

Table 3: cross tabulation of age, education and gender

The internal consistency reliability was measured using the Cronbach’s Alpha Based on standardised items, and as shown in table 4, the result was 0.913. According to Cronbach and Shavelson (2004), the closer the value is to 1, the more desirable the instrument used is, with the minimum reliability requirement being 0.7 or higher. In this case, therefore, the score of 0.913 shows that the instrument used is reliable.

Further, the study determined the inter-item correlation between the 64 items used in the study, and by pairing them into groups of 8, the mean of the inter-item correlation was determined as 0.166, with a variance of 0.43 as shown in table 4. This indicates a weaker relationship between the items. In the same vein, the scale mean if a particular item was deleted was also carried out, and the data shows that the average mean is 56.1, whereas the Cronbach’s alpha ranges between 9.17 and 9.09.  This data shows that the weighting per item is relatively even, and no single item carries the entire weight of the experiment.

Table 4: summary item statistics

As regards the correlations, the data shows that the Pearson Correlation coefficient between age and CCT items is .017 significant at the 0.05 level. This shows that there is a weak positive relation between age and the CCT items.

Table 5: Pearson’s Correlation

On the other hand, from the Levene’s Test for Equality of Variances, it was determined that sig. value (0.85) is more than 0.05, which, according to Levene (1960), indicates a small variance between the males and females. Moreover, the Sig. (2-tailed) value (0.642) was greater than 0.5, which indicates that there is no statistically significant difference between the CCT score between males and females.

Table 6: Independent sample T-test for gender and CCT

Similarly, from the NPar test conducted using the Kruskal-Wallis Test, it was observed that there is a significant difference in CCT scores for each education category, x²(2) = 4.835, p = 0.089, with a mean rank CCT_SUM score of 16.64, 25.95 and 30.76 for primary, secondary, and higher education respectively.     

In summary, the data presented above shows that there is a correlation weak but positive relation between age and CCT_SUM, but there is a no statistically significant relation between gender and CCT_SUM. Comparatively, there is a significant difference in the result for different education levels, which implies that education has a relation to the CCT results.

Discussions

From the data presented in the section above, this study confirms a number of observations made in previous studies. Firstly, the findings resonate what Rogers et al. (2006) found in their studies. It also mirrors the results given by Ganguli et al (2010), who conducted a study on the effects of age and education on the cognitive test battery. These researchers focused on individuals aged 65 years upwards, and living within the urban community of USA. They concluded that less education and old age are more likely to lead to poorer neuropsychological test performance among normal adults. While the present study used a majorly younger population, the finding mirrors their findings to a significant extent, thereby indicating the need to carry out further studies in this respect.

Additionally, the findings agree with the conclusions of Acevedo et al. (2007), Ganguli et al. (1991) and other scholars who confirmed the relationship between education, age and neuropsychological tests. On the other hand, the findings on gender indicate that there is no link with the test results, showing that females are equally as likely as males to be affected (DeVellis, 2003; Borsboom, 2005; Warne, Yoon and Price, 2014; Hernández-Orallo, 2013; Wilson, 2004). This finding is also in agreement with that of Acevedo et al. (2007), both in normal people and in those suffering from dementia.

These findings should be of value to researchers and clinicians as an addition to the on-going literature on the subject. However, the limitations of scope and scale of this study are not lost to the researcher, especially the minimal amount of data used, and the use of convenience sampling technique, which increases the potential for biases. Because of the limited scope used, these results cannot be generalised to represent mainstream population and it is advisable that they only be used within the context of the population from which the participants were drawn. Nevertheless, by supporting and criticising these findings using secondary findings from peer-reviewed sources, the extent of validity and reliability is increased.        

References

Acevedo A, Loewenstein DA, Agron J, Duara R. (2007). Influence of sociodemographic variables on neuropsychological test performance in Spanish-speaking older adults. Journal of Clinical & Experimental Neuropsychology: Official Journal of the International Neuropsychological Society.;29(5):530–544

Bonett, D.G. (2003). “Sample size requirements for comparing two alpha reliability coefficients”. Applied Psychological Measurement (27): 72–74.

Bonett, D.G. (2010). “Varying coefficient meta-analytic methods for alpha reliability”. Psychological Methods 15: 368–385.

Borsboom, D.(2005). Measuring the Mind: Conceptual Issues in Contemporary Psychometrics. Cambridge: Cambridge University Press.

Bozeat, S., Lambon Ralph, M.A., Patterson, K., Garrard, P., & Hodges, J. R. (2000). Non-verbal semantic impairment in semantic dementia. Neuropsychologia, 38, 1207–1215.

Cronbach, Lee J.; Shavelson, Richard J. (2004). “My Current Thoughts on Coefficient Alpha and Successor Procedures”.Educational and Psychological Measurement 64 (3): 391–418.

DeVellis, R.F (2003). Scale Development: Theory and Applications (2nd ed.). London: Sage Publications.

Ganguli, M., Snitz, B.E., Lee, C.W., Vanderbilt, J., Saxon, J.A., and Chang, C.H. (2010) Age and education effects and norms on a cognitive test battery from a population-based cohort: The Monongahela –Youghiogheny Healthy Aging Team (MYHAT). Aging Ment Health. 2010 Jan; 14(1): 100–107.

Grundman, M., Petersen, R. C., Ferris, S. H., Thomas, R. G., Aisen, P. S., Bennett, D. A., et al. (2004). Mild cognitive impairment can be distinguished from Alzheimer disease and normal aging for clinical trials. Archives of Neurology; 61, 59–66.

Hernández-Orallo, J.; Dowe, D.L.; Hernández-Lloreda, M.V. (2013). Universal Psychometrics: Measuring Cognitive Abilities in the Machine Kingdom. Cognitive Systems Research.

Levene, H. (1960). “Robust tests for equality of variances”.et al. Contributions to Probability and Statistics: Essays in Honor of Harold Hotelling. Stanford University Press. 278–292.

Rogers, T. T., Ivanoiu, A., Patterson, K., & Hodges, J. R. (2006). Semantic memory in Alzheimer’s disease and the frontotemporal dementias: A longitudinal study of 236 patients. Neuropsychology, 20(3), 319–335.

Rogers, T. T., Lambon Ralph, M. A., Hodges, J. R., & Patterson, K. (2003). Object recognition under semantic impairment: The effects of conceptual regularities on perceptual decisions. Language and Cognitive Processes, 18, 625–662.

Warne, R. T., Yoon, M., & Price, C. J. (2014). Exploring the various interpretations of “test bias.” Cultural Diversity and Ethnic Minority Psychology, 20, 570-582. 

Wilson, M. (2004). Constructing Measures: An Item Response Modeling Approach. Mahwah. NJ: Lawrence Erlbaum.