Research Methodology: Design, Population, Instruments, Data Collection, And Analysis

The Importance of Employing Effective Research Methodologies

Discuss about the Increasing Value and Reducing Waste in Design.

This chapter will give more details of the research methodology that was used in this study. Some of the subsections contained in this chapter includes; the research design, targeted population, research instruments used, methods for data collection and lastly the data analysis. Capturing the aforementioned subsections will build up the validity of and acceptability of this research study.

This is the general strategy that the researchers can choose to combine various components in the study in a logical and coherent way, thus ensuring for appropriate and effective research problem address comprising of data collection, measurement and ultimately the analysis of data (Ioannidis et al, 2014). This step is important for it reduces the ambiguity in the effectiveness of the research Vogt, W. P., Gardner, D. C., & Haeffele, L. M. (2012). As it is the case in social science research, having relevant information that satisfies the research problem is important for the accurate test of the program and the test of theory drawing meaning from observable related events. In almost all the instances, it was always vital to carry out critical evaluation and examination of the methodologies to be used in any of the research studies in order to make them reliable and thus generate valid results that will later be scrutinized for the work done Smith (2015). Longitudinal design and some of the control groups were used in the research to examine the participation results in the data collection process Twisk (2013). In this research study, the descriptive survey design was used with the aim of assessing the legalization of marijuana attitudes for either medical or recreational use. The process of qualitative and quantitative data collection was employed by the researcher where the data of interest were collected from the targeted group. The types of data collected were with the intention of meeting the measures for use in the research.

Population is a collection of elements or objects aimed by the topic of study from where sample of suitable size is chosen and data drawn (Wohlfahrt et al, 2016). A sample is a proportion drawn from the population and sample size is the number of elements or objects of study (Yabroff et al, 2016). A population of 10,000 respondents were aimed at for the collection of data for this study for which people of varied cultural ethnic background were to be participants. From the population, a sample of 164 participants was used where they participated in the data collection of data by filling the questionnaires voluntarily. Out of which 43.9% of the participants were male while the remaining 55% were female. Legalizing marijuana was the topic of study questions where they were to give their responses in regards to the questions given. The sample size was relatively good for use since it covered almost thirty percent of the targeted population Ward, B. W., Schiller, J. S., & Goodman, R. A. (2014). Using a relatively small sample size made it easier and cost effective in the collection of data, time was as well saved in the process but all these were done with the aim of achieving accurate results Aggarwal (2015). In meeting the research target, it is postulated by other researchers that experienced participants with the subject of study will result to collection of detailed information that would be helpful towards making the study a success (Ambrus et al, 2012). Responses thereafter used to give the results that were taken for use in the data analysis part for the research study results.

Research Design and Population

The measurement tools used in the research that are designed particularly for the collection of data concerning the topic of interest Marek, T., Schaufeli, W. B., & Maslach, C. (2017). There are various research instruments but the choice for which one to use depended on the purpose aimed at by the study Bryman (2017). In this study, the researcher used the well-structured questionnaires. Questionnaires are known and more popular for their purpose in the data collection process (De Jong et al, 2013). In response to that, questionnaires always have the propensity of allowing the participants with free time to express their feelings concerning the subject of study De Vaus (2013). Different types of questions structure were used in the questionnaire including open and closed ended questions. Open ended questions were used to let the participants to freely give response of their choice concerning the question’s requirement Oberjé, E. J., Dima, A. L., Pijnappel, F. J., Prins, J. M., & de Bruin, M. (2015). Closed ended questions are in most of the cases preferred by researchers for their ease in the times of analysis since they can easily be coded Kazi, A. M., & Khalid, W. (2012). Moreover, they as well gave the participants easy time especially when responding to the questions they never had much idea about. Being that the Likert scale was used in the questions provided in the questionnaires for the attitude of the participants towards legalization of marijuana, Cronbach’s alpha was used in reliability test Bonett, D. G., & Wright, T. A. (2015). The research instrument that was used in the research had higher internal consistency thus reliability of the questionnaires since the Cronbach’s alpha was 0.805. In case where the instrument used poses to have very smaller values or too high values, the instrument will show not to be reliable (Malmstrom et al, 2016). In connection to that therefore, those values are to be removed for the improvement of the reliability and psychometric properties in general Malmstrom, T. K., Miller, D. K., Simonsick, E. M., Ferrucci, L., & Morley, J. E. (2016). All the selected 14 items belonged to the same scale as by the test conducted where corrected item total correlations were greater than (0.30) for all items. When the number of items were reduced to 7, the Cronbach’s alpha increased to 0.840 still confirming for the internal consistency of the questionnaires used, means and standard deviations in the item statistic table further confirmed this since they were too close to one another and finally, the corrected item correlations for the reduced number of items was still greater than 0.30 to mean that all the items were still falling on the same scale.

Research Instruments Used

Obtaining data from the participants, did take some of the procedures among which the issuance of the questionnaires to the participants was preferred by the researcher. The printed questionnaires were distributed by the researcher to all the 164 participants under his administration. The participants were required to respond to all the provided questions on the questionnaire by either indicating a tick on their choices or providing explanations where necessary.

Primary data were collected in the form of responses from the participants in the questionnaires where they were entered in the Statistical Package for Social Sciences (SPSS) for data analysis. Organization and cleaning of the data was done where missing data were dealt with for the validity of the study results Wang, L., Da Xu, L., Bi, Z., & Xu, Y. (2014). KMO and Bartlett’s test of sphericity was used in order to decide on which analysis process to be conducted. In response to the test conducted (KMO and Bartlett’s Test), the test’s significance value was less than 0.05. This indicated and permitted the researcher to use the factor analysis. In the extraction of factors to be used, principal component analysis was employed by the researcher. Furthermore, in order to determine the orthogonally correlated factors and come up with correlated factors, direct oblimin rotation was used by the researcher and to determine the number of available factors used in the study, Eigenvalues were used.

Table 1: Demographic information

Variables	Statements	Frequency	Percent (%)
Gender	male	72	43.9
female	91	55.5
Education	no formal education	1	.6
primary level	2	1.2
lower secondary level	10	6.1
upper secondary level	79	48.2
technical or vocational qualification	3	1.8
both upper secondary or technical or vocational qualification	12	7.3
third level	57	34.8
College	Yes	87	53.0
No	76	46.3
Ethnic	Irish	145	88.4
Irish  traveler	3	1.8
other white background	13	7.9
other	3	1.8
Religion	Yes	60	36.6
No	103	62.8
Total	163	99.4
System	1	.6

The number of male participants was represented by 43.9% whereas that of female participants was 55.5%.  0.6% of the participants had no formal education, 1.2% of the participants got primary level education, 6.1% had lower secondary level of education, 48.2% of the participants reached upper secondary level, 1.8% represented those who had technical or vocational qualification,  7.3% represented both upper secondary or technical or vocational qualification and 34.8% were third level graduates. 53% of the participants attended college while 46.3% did not. 88.4% represented the respondents whose ethnic was Irish, 1.8% represented Irish traveler, 7.9% represented respondents with other white backgrounds and the remaining 1.8% represented the respondents of other ethnicities. 36.6% represented the respondents with religion while 62.8 represented those who responded to never had religion.

Table 2: KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.	.904
Bartlett’s Test of Sphericity	Approx. Chi-Square	4691.424
df	1176
Sig.	.000

The measure of sampling adequacy from the test was found to be .904 indicating that the sample was adequate. Further, the significance test showed that factor analysis could be used as the statistical significance was less than 0.05.

Table 3: Smoke
	Frequency	Percent	Valid Percent	Cumulative Percent
	yes	37	22.6	22.7	22.7
no	126	76.8	77.3	100.0
Total	163	99.4	100.0
Missing	System	1	.6
Total	164	100.0

Data Collection Methods

Out of the participants who participated in filling the questionnaires, 22.6% of them were smokers of marijuana while the rest represented by 76.8% were not smoking marijuana.

Hypothesis

Hypothesis was tested to check for statistical significance in the correlation that existed between age and gender for their attitudes towards legalization of marijuana

H₀: There is no statistical significance in correlation between age and gender for their attitudes towards legalization of marijuana

H₁: There is statistical significance in correlation between age and gender for their attitudes towards legalization of marijuana.

Table 4: Correlation between age and gender
	Age	Gender
Age	Pearson Correlation	1	-.032
Sig. (2-tailed)		.685
N	164	163
Gender	Pearson Correlation	-.032	1
Sig. (2-tailed)	.685
N	163	163

There was a weak negative correlation of (-.032) between age of participants and their genders for their attitude towards legalization of marijuana. The statistical significance being (.685) which was far much greater than .05, it therefore showed that no statistical significance existed between the two variables i.e. age and gender. We therefore fail to reject the null hypothesis that there was no statistical significance in correlation between age and gender for their attitudes towards legalization of marijuana.

Table 5: Total Variance Explained
Component	Initial Eigenvalues	Extraction Sums of Squared Loadings	Rotation Sums of Squared Loadingsa
Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %	Total
1	17.347	35.402	35.402	17.347	35.402	35.402	7.727
2	2.350	4.796	40.198	2.350	4.796	40.198	3.258
3	1.941	3.961	44.159	1.941	3.961	44.159	4.554
4	1.801	3.675	47.834	1.801	3.675	47.834	3.422
5	1.556	3.175	51.009	1.556	3.175	51.009	1.961
6	1.499	3.059	54.069	1.499	3.059	54.069	6.765
7	1.361	2.777	56.846	1.361	2.777	56.846	3.000
8	1.307	2.667	59.512	1.307	2.667	59.512	9.570
9	1.201	2.451	61.964	1.201	2.451	61.964	5.568
10	1.107	2.260	64.223	1.107	2.260	64.223	4.492
11	1.096	2.237	66.460	1.096	2.237	66.460	6.325
12	1.047	2.136	68.596	1.047	2.136	68.596	8.046
13	.994	2.028	70.624
14	.927	1.892	72.517
15	.813	1.659	74.176
16	.772	1.575	75.751
17	.760	1.552	77.303
18	.734	1.498	78.800
19	.694	1.417	80.218
20	.625	1.276	81.494
21	.619	1.263	82.757
22	.546	1.114	83.871
23	.536	1.094	84.965
24	.502	1.024	85.989
25	.481	.981	86.970
26	.471	.961	87.931
27	.459	.937	88.868
28	.425	.868	89.736
29	.417	.851	90.587
30	.383	.781	91.368
31	.375	.765	92.134
32	.348	.711	92.844
33	.334	.681	93.526
34	.322	.658	94.184
35	.311	.635	94.819
36	.278	.567	95.386
37	.263	.538	95.924
38	.253	.517	96.441
39	.242	.494	96.934
40	.207	.422	97.357
41	.196	.400	97.756
42	.183	.373	98.130
43	.164	.335	98.464
44	.155	.317	98.781
45	.147	.299	99.080
46	.132	.269	99.349
47	.120	.245	99.594
48	.108	.220	99.814
49	.091	.186	100.000
Extraction Method: Principal Component Analysis.
a. When components are correlated, sums of squared loadings cannot be added to obtain a total variance.

From the total variance explained table, total of 49 factors were obtained as in the component column. The Eigenvalues column showed the variances of the used 49 factors in this study. Standardization of each variable made them have variances of 1 and the total variance given showed the number of variables used in the study analysis (i.e. 49). In the total column, the first factor was seen accounting for the largest number of variance of 17.347 followed by second factor and the trend continues decreasing until the last factor with the least variance of .091. The percentages of variances of each factor used were labelled in the (% of variance column). The number of retained factors from the factor analysis were 12 as shown by the extraction sums of squared loadings columns upon the request of their retention. There are 12 rows for each one of the factors retained in the table 5 above.

From the factor analysis, the correlation of attitudes of those who felt that should marijuana be made legal, it should be with doctor’s prescription and those who felt that people who smoke marijuana everyday could not live fully functional life was a negative correlation of (-.377). A positive correlation of (.191) was posted for correlation between those who felt that legalizing marijuana would benefit the economy and those who felt that people who smoke marijuana everyday could not live a fully functional life. Some respondents thought that marijuana should be made available in the pharmacies posted a positive correlation of (.219) with those who believed that people who smoke marijuana could not live fully functional lives.

Data Analysis

The scree plot graph consisted of the Eigenvalues against the number of components, the graph steeply reduces from the first factor values and decreases in steps until a point where it seemingly looks flat towards the last factors. Showing the variance changes in the factors.

Factors extraction was done using the principal component analysis and the KMO and Bartlett’s test of sphericity 0.906 confirming for the facto analysis. The variables used were computed and they were named Sub1_T1 and Sub2_T2. The Cronbach’s alpha was 0.840 for the test of internal consistency and reliability of the subscales used and thus they were reliable as per the test. This was further confirmed by the closeness of their means and standard deviations. Mean, standard deviation, minimum and maximum values for the variable were calculated using descriptive statistics analysis.

From the descriptive statistics, the minimum values for both variables was 1 whereas their maximum values were 5, their means ranged from 1.92 to 4.32 with the majority having the mean of above 2. Standard deviations of the ranged from .974 to 1.372. The standard deviation values are closely distributed around the mean thus confirms the spread and concentration of the values around the mean. Descriptive statistics table is in the appendices.

Hypothesis

Among various tests of hypothesis, the researcher used correlation for the test of gender and their attitudes towards legalization of marijuana.

H₀:  There is no significant difference between gender and their attitude towards legalization of marijuana

H₁: There is significant difference between gender and their attitude towards legalization of marijuana

Table 6: Correlations
	Sub 1_T1	Sub 1_T2
Sub 1_T1	Pearson Correlation	1	.898**
Sig. (2-tailed)		.000
N	14	12
Sub 1_T2	Pearson Correlation	.898**	1
Sig. (2-tailed)	.000
N	12	17
**. Correlation is significant at the 0.01 level (2-tailed).

Pearson correlation value (.898) between Sub 1_T1 and Sub 1_T2 showed that there was a strong correlation between gender of a person and their attitude towards legalization of marijuana.

Further this test result was confirmed by the significance test. With the significance value of less than .05, we reject the null hypothesis and confirm the alternative hypothesis true i.e. that there was significant difference between gender and their attitudes towards legalization of marijuana. 

Table 7: Correlations
	Sub 2_T1	Sub 2_T2
Sub 2_T1	Pearson Correlation	1	.562
Sig. (2-tailed)		.057
N	14	12
Sub 2_T2	Pearson Correlation	.562	1
Sig. (2-tailed)	.057
N	12	17

When the variables were interchanges and retested, the Pearson correlation value was .562 which confirmed the coexistence of strong relationship between gender and their attitudes towards legalization of marijuana. But that was never the case as the statistical significance test was .057 which is greater than .05 thus we fail to reject the null hypothesis and conclude that there was no significant difference between gender and their attitude towards legalization of marijuana.

References

Ambrus, G. G., Al-Moyed, H., Chaieb, L., Sarp, L., Antal, A., & Paulus, W. (2012). The fade-in–short stimulation–fade out approach to sham tDCS–reliable at 1 mA for naive and experienced subjects, but not investigators. Brain Stimulation: Basic, Translational, and Clinical Research in Neuromodulation, 5(4), 499-504.

Aggarwal, C. C. (2015). Outlier analysis. In Data mining (pp. 237-263). Springer, Cham.

Bonett, D. G., & Wright, T. A. (2015). Cronbach’s alpha reliability: Interval estimation, hypothesis testing, and sample size planning. Journal of Organizational Behavior, 36(1), 3-15.

Bryman, A. (2017). Quantitative and qualitative research: further reflections on their integration. In Mixing methods: Qualitative and quantitative research (pp. 57-78). Routledge.

De Jong, E., Visscher, T. L. S., HiraSing, R. A., Heymans, M. W., Seidell, J. C., & Renders, C. M. (2013). Association between TV viewing, computer use and overweight, determinants and competing activities of screen time in 4-to 13-year-old children. International journal of obesity, 37(1), 47.

De Vaus, D., & de Vaus, D. (2013). Surveys in social research. Routledge.

Ioannidis, J. P., Greenland, S., Hlatky, M. A., Khoury, M. J., Macleod, M. R., Moher, D., … & Tibshirani, R. (2014). Increasing value and reducing waste in research design, conduct, and analysis. The Lancet, 383(9912), 166-175.

Kazi, A. M., & Khalid, W. (2012). Questionnaire designing and validation. Journal of the Pakistan Medical Association, 62(5), 514.

Malmstrom, T. K., Miller, D. K., Simonsick, E. M., Ferrucci, L., & Morley, J. E. (2016). SARC?F: a symptom score to predict persons with sarcopenia at risk for poor functional outcomes. Journal of cachexia, sarcopenia and muscle, 7(1), 28-36.

Marek, T., Schaufeli, W. B., & Maslach, C. (2017). Professional burnout: Recent developments in theory and research. Routledge.

Oberjé, E. J., Dima, A. L., Pijnappel, F. J., Prins, J. M., & de Bruin, M. (2015). Assessing treatment-as-usual provided to control groups in adherence trials: exploring the use of an open-ended questionnaire for identifying behaviour change techniques. Psychology & health, 30(8), 897-910.

Smith, J. A. (Ed.). (2015). Qualitative psychology: A practical guide to research methods. Sage.

Twisk, J. W. (2013). Applied longitudinal data analysis for epidemiology: a practical guide. Cambridge University Press.

Vogt, W. P., Gardner, D. C., & Haeffele, L. M. (2012). When to use what research design. Guilford Press.

Wang, L., Da Xu, L., Bi, Z., & Xu, Y. (2014). Data cleaning for RFID and WSN integration. IEEE transactions on industrial informatics, 10(1), 408-418.

Ward, B. W., Schiller, J. S., & Goodman, R. A. (2014). Peer reviewed: Multiple chronic conditions among us adults: A 2012 update. Preventing chronic disease, 11.

Wohlfahrt, P., Cífková, R., Movsisyan, N., Kunzová, Š., Lešovský, J., Homolka, M., … & Sochor, O. (2016). Threshold for diagnosing hypertension by automated office blood pressure using random sample population data. Journal of hypertension, 34(11), 2180-2186.

Yabroff, K. R., Dowling, E. C., Guy Jr, G. P., Banegas, M. P., Davidoff, A., Han, X., … & Kent, E. E. (2016). Financial hardship associated with cancer in the United States: findings from a population-based sample of adult cancer survivors. Journal of Clinical Oncology, 34(3), 259.