Advertisements

Wise Running

Train smart, eat well, and enjoy the run.

Choosing Research Methodology

This page is still under construction…

 

The most important decision after crafting your research question(s) is choosing appropriate research methodologies.  The methodology must be consistent with the conceptual framework, the theoretical framework, and the wording of the research question.

Quantitative of Qualitative

The first part of that process is determining whether your question is more consistent with qualitative or quantitative methods.  Quantitative questions ask about significant differences, whether a relationship exists between two or more variables, and the extent to which the change in a dependent variable (outcome) is dependent on any one independent variable.  More complex test can answer similar questions with multiple IVs, multiple DVs, and control for the affect of confounding covariates.

While quantitative methodology cannot accomplish any of those tasks, it can do many things that quantitative research is unable to offer.  Qualitative research can provide the non-quantifiable insights that allow quantitative questions to be asked.  In that respect, qualitative research can identify and define the variables that can then be defined in a quantitative study.  The deeper contribution of qualitative research is found in its ability to analyze non-quantitative data in an organized and logically consistent way.

Dr. Taylor’s example:

For my dissertation, I studied the nature and extent of collegial interaction and the contextual catalysts that support or deter such collegial learning in efforts to support teacher and student learning.  This was clearly a qualitative study.

However a colleague and I followed up this qualitative study by creating taxonomy (rubric) for evaluating the extent of collegial interaction and contextual catalysts.  It was a tool to quantitatively study that which had been considered to be strictly qualitative.

The first study explored interaction in schools through observation, interviews and artifacts.  It helped define the key terms within the quantitative rubric used in the follow-up.

Choosing Quantitative Statistical Tools

In most cases, quantitative methodology can be determined by the specific language of the research question.

Below is a table with the most commonly used tests in the doc program at CNU:

Step 1:  Identify the variables listed in your research question.

Step 2:  For each variable, label it as dependent (outcome) or independent (input).

Step 3:  For each variable, identify the data type (dichotomous, categorical/nominal, ratio, interval, continuous

Step 4: Examine the research question to determine if the goal is to identify the purpose (see last column of the table).

Step 5:  Find the row in the table that matches your research question in all of columns 2-6 in the table.

Step 6:  The quant test/tool you need is named in the first column of that row.

best stat table ever

Step 7:  Identify the population to which you want to generalize the results.  Then choose a sample that is representative of the larger population.  Make sure you choose a larger sample.  20 participants is a tiny study.  The farther you go over 20, the more trustworthy the results.  

Step 8:  Control for extraneous or confounding variables.  Many things impact the outcomes in addition to the independent variables you have found.  You can reduce the effect of these variables by sampling or by using covariates.  Matching the individual attributes of participants in the control and experimental group is a good way to remove the affect of the variable.  For instance, if gender may play a role in the outcome it is best to choose an equal number of males and females in the control and experimental groups.  The other way to consider an extraneous or confounding variable is by using it as a covariate in such texts as ANCOVA, MANCOVA, and multiple regression.

Step 9:  Choose the number of tails on the test and the confidence level.  Typically, we want to ensure the results are 95% certainty.  This would require the p-value to be less than or equal to 0.05.  The tests are nearly always run as two-tailed because it is possible for results to go the opposite of what you expected.  Only in rare instances can you make the case that it could only have results on one end of the spectrum.

Step 10:  Collect the data in ways that provide both validity and reliability.  The procedures and the instruments much be both logical and proven and supported in previous work.  If the validity of the instruments was not determined in a prior study, you must consult with experts in the field as well as running a pilot study to ensure both validity and reliability of the instrument are established.

Choosing Qualitative Tools

Qualitative research is similar in that the specific structure and intent of the research questions can determine quite a bit of the methodology.  However, there tends to be more flexibility in the data collection methods in qualitative research.  Determining the specific methods may lean even more on the theoretical and conceptual frameworks.

Qualitative research depends on qualitative data which includes anything that is not quantified and possibly some descriptive quantitative statistics as well.

The

 

This page is still under construction…

 

 

 

 

Advertisements
%d bloggers like this: