Research Methods in Education - Case Study Example

Chase and Drummer’s Use of Quantitative Data Introduction: Use of statistics and quantitative techniques in research work in social science is something well established. The main driving force behind going for a quantitative research work is that the findings can be quantified and therefore can be compared with those of the other contemporary researcher works and even with the historical work with similar nature to see if there is any trend in the findings with time. Important statistical aspects of any quantitative research work are – Data Collection Instrument or Questionnaire, Nature of Data, Sample, Analysis of Data and Drawing conclusions. In this paper, use of Quantitative Data by Chase M. A. and Dummer G. M. will be briefly discussed and more importantly it will be analyzed whether or not they have conducted experiment appropriate for answering the question they wanted to answer, whether or not they have presented the data in an appropriate manner and also how the experiment could have be improved further. Data Collection Instrument: Data collection instrument in the experiment conducted by Chase and Dummer was a questionnaire developed and used by researchers like Buchanan et al (1976), Coleman (1976) and Feltz (1978). Additionally they have also collected additional data of pertaining to the demography of the subjects (students) like gender, age, grade, ethnic background etc. They have also included some more questions in addition to what has been there in the questionnaire developed and used by the previous co-workers. Whenever, the data is collected using a questionnaire it is important to know how reliable was the instrument for collecting the data it is intended to collect. What it means if the same questionnaire is presented to the same group of subjects after a gap of reasonable time span so that he has forgotten what he had reported earlier, then how much similarity will the two set of response will show. To test reliability of the data collection instrument (Questionnaire), researchers usually conduct “Test – Retest Analysis” using the questionnaire with a small set of respondents say 20 or 25 subjects at a gap of two weeks or so and then they compare the two set of response to analyze the reliability of the Data Collection Instrument (Questionnaire) by using Kappa Statistics K. This effectively tests whether or not there is internal consistency in the data collection instrument (Cohen et al 2007). However, Chase and Dummer have not conducted the reliability test on their data collection instrument. This could have been due to the fact that they had simply taken the questionnaire from previous researchers, who might have confirmed the reliability of the questionnaire. But even then it was duty of Chase and Dummer to talk about the reliability of the data collection instrument they used in this study. This is a very important improvement that is required in this study. Types of Data In any quantitative study one measures and analyzes data pertaining to different types of variables like Continuous, Interval, Discrete, ordinal, nominal categorical etc. Some of the variables with different attributes, which have been referred to are briefly described below. (i) Variables with Interval Properties – The variables of this nature referred to in this paper by Chase and Dummer are Age of the subjects, Grades of the subjects and family income of the subjects. (ii) Ordinal Variables – Ordinal variables refer to some kind of order – increasing / decreasing, larger / smaller etc. Some ordinal variables measured in this study by Chase and Dummer are - Age of the subjects, Grades of the subjects and family income of the subjects as these variables show the relative order like standard five is higher than standard 4 and so on. (iii) Categorical or Nominal Variables – These variables refer to some category. Like gender of subjects where ‘1’ stands for Male and ‘2’ stands for Female. Here numerical value of the variable is of no significance. One could have even chosen ‘10’ to represent males and ‘20’ to represent females, without any implication on the results. Similarly response of the subjects against different question was ‘1’, ‘2’, ‘3’ and ‘4’. Here also the numerical values are only categorical. However, in this case the numerical value has some kind of order like ‘1’ one stands for strongly agree and ‘4’ for strongly disagree and the intermediate values refer to the intermediate stages of agreement. Therefore, this variable is ordinal as well. There are some variables like gender and grade which are properties of a group and not that of only an individual. Presentation of Data Presentation of data in any report of quantitative study is very important. A nicely presented data goes deep into the mind of a reader and therefore has better impact and therefore, generates better response. In this study by Chase and Dummer, the data and results were presented only in tabular form and so much information is there in the tables that a reader gets lost in that and s/he has to put lot of strain in extracting information from the tables. This is certainly very boring presentation of the data and results. It will certainly fail to evoke any response from a common reader and those working in this area will also find it straining to comprehend, what is presented in these tables. It is said that a picture is worth thousand words, however, it appears Chase and Dummer have not heard this famous proverb. They should have presented the results using suitable diagrams as well. For example the results reported in Table 1, should have been presented using a bar chart with error marks. Similarly the results reported in table 2 and table 3, should have been presented using either a bar chart or a pie chart. A pictorial representation of the results besides the tabular representation should have been more attractive for a common reader and professionals alike. The measure of central tendency i.e. mean and that of the dispersion i.e. standard deviation has been simply reported as numbers. In pictorial representation while mean is represented by a point, the standard deviation is represented by an error bar of the length equal to the magnitude of the standard deviation on the either side of the point. Such a representation is very good in conveying to a reader that there is some uncertainty in the measurement of the variable and the extent of the uncertainty was within the error bar. What was most interesting for me is the finding that appearance is the most important for girls. Though it is something that is widely accepted, but I never anticipated it comes to them from such an early stage like 4th and 5th standard. If so is the case, then this has to do more with the biological differences between the two sexes than the social mind set about the two genders. Therefore, I am not concerned like Chase and Dummer that girls give more importance to their appearance. If they do so and they are so made by the Nature, then what is harm in that? I cannot understand why only achievements in education, sports and profession should be the yardstick to measure good or bad? Humans are not machines and therefore, soft attributes like beauty, tenderness etc. are equally important for holistic development of personality and therefore, there is no harm in girls focusing more on their appearance. What was least interesting for me was the preference of the subjects towards grades. In my opinion the students were still very early in their academic career and at this stage it hardly matters, how much they care about the grades. I agree only partially (may be about 60%) that the data collected allowed the researchers to effectively answer the questions they intended to answer. In my opinion instead of going for categorical response like ‘1’, ‘2’, ‘3’ and ‘4’ for against their questions, they should have been more innovative and they should have rather used a continuous response. For example they should have given 100 points to each of the students to be distributed among the four responses. This should have made the response a continuous variable in stead of a categorical variable. The benefit is that by doing so they would have got a response in which liking or disliking of the subjects toward different activities could have been compared. Like a student likes sport to the extent of 80%, academics just 5%, appearance just 10% and mney just 5%. Here there is a quantitative relationship between his like for different things, an information, which can never be extracted from responses like ‘1’, ‘2’, ‘3’ and ‘4’. This would not have required any additional data, just nature of the data would have got changed and a lot more information could have been derived subsequently. This is the idea I would like to present at this point in the course. References: Buchanan, H. T., Blankenbaker, J and Cotton D. “Academic and athletic ability as popularity factors in school settings”. L. G. Katz (Ed.), Current topics in early childhood education, pp. 33 – 61, Norwood, NJ:Ablex, 1976 Cohen L., Manion L., Morrison K., “Research Methods in Education”, 6th Ed., Routledge, 2007 Coleman J. S. “The adolescent society: The social life of the teenager and its impact on education”. New York: Free Press, 1976. Feltz, D. L. “Athletics in the system of female adolescents”. Review of Sports and Leisure, 1 (1978) 98 – 108. Buchanan et al (1976), Coleman (1976), Feltz (1978) and Thirer and Wright (1985) Read More