Inscrivez-vous ou connectez-vous pour rejoindre votre communauté professionnelle.
Does the type of data collected for such research work restrict your choices in such particular of data analysis method, rather than others? Whatever your job role, I can accept any opinion you suggests in this issue. thanks
Two basic research types
Practical : consists of the empirical study of the topic under research and hands on approach.
Theoretical : researching through archives of public libraries, court rooms and published academic journals.
In Practical Research the methods used will be more pratical and based on observations and hands on approach such as interviews, questionnaires, surveys, etc.
In theoretical approach it is mostly based on past data. like archives, journals as i have already mentioned
So i feel that there is a correlation between the type of Research you undertake and the method that will be used.
Of course. The research type;ie....quantitative or qualitative the research design is important and connected to complete study
no there is no correlation between Research Type and Method type used in data analysis.
The basic steps of any quantitative research are as follows:
1. Define your research question
2. Design appropriate study
3. Collect the data
4. Analyze the data
5. Draw conclusions from the data
Step4, which is what we are talking about , does not really depend on Steps1 &2 but actually relies heavily on Step3- the Nature of variables involved and the type of data collected. Let me illustrate the point.
Suppose you are interested to know whether there is any relationship between exercise and energy level among healthy adults, your research question could be as follows:
Is exercising correlated with higher energy levels among healthy adults?
OR
Does exercising cause higher energy levels among healthy adults?
To answer the first research question you have to design an Observational or correlational study. To do this you would typically do a cross-sectional study in which you would randomly select healthy adults from your population of interest and classify them into two or more groups based on their exercise status. Then you would measure the energy level of each participant and compare the mean energy levels of the groups using a t-test or ANOVA to check for significant differences.
However if your research question is the second one then you would design an experimental or causal study in which you would enroll participants and randomly assign them to two or more groups and give them different instructions regarding exercise. Then after some time you would measure the energy level of each participant and compare the mean energy levels of the groups using a t-test or ANOVA to check for significant differences.
Please note that whether your research is observational or experimental , the method of data analysis is still the same, i.e, you use a t-test or ANOVA in both cases. The reason for this is that in both the above studies the outcome variable is Energy level ( which is quantitative)
So , basically data analysis methods depend on the following3 aspects:
1. Is the outcome variable quantitative ( eg. age, cognitive function) or binary/ categorical( eg. pregnant yes/no)?
2. Are the observations independent or correlated?
3. Are the assumptions of normality and homoscedasticity violated?
You could use the following guidelines:
1. If outcome variable is continuous and
a. the observations are independent- t-test, ANOVA,Linear correlation/regression
b. the observations are correlated- Paired t-test, Repeated Measures ANOVA
c. the assuptions are violated- Wilcoxon sign -rank test, Wilcoxon rank-sum test, Kruskal-Wallis test, Spearman rank correlation
2. If outcome variable is Binary/categorical and
a. observations are independent- Risk difference/ relative risks, Chi-square test, Logistic regression
b. observations are correlated- McNemar's test, Conditional logistic regression.
c. assumptions are violated- Fisher's exact test, McNemar's exact test.
Yes, there is a correlation between the research type and the methodology and methods of data analysis. For example, quantitative research vs qualitative, each has different methods and data analysis technique. In addition, if you are conducting empirical study, you have to make choice between the parametric and parametric techniques of data analysis. We have to use the parametric techniques if the dependent variable is dichotomous (0,1; Yes, No)