Understanding quantitative research methods

Strengths of quantitative, qualitative, and mixed-method research

Any theory cannot be viable unless it has a viable empirical basis, which is why the methodology of the studies currently requires empirical validation using both quantitative and qualitative methods.

Quantitative and qualitative methods are used to support the analysis of an economic (inflation, unemployment, etc.), pathological (evolution of COVID19, etc.), biological (effectiveness of a treatment, for example), etc. phenomenon, as well as to relate the phenomenon to other phenomena.

But to apply these methods, the phenomenon must be measured, so its evolution must be translated into figures.

Illustration by an example: 

Theory: Chloroquine is effective against coronavirus. 

Empirical validation: take a sample of patients who tested positive (for example 100 patients), subject 50 of them to chloroquine treatment.

The phenomenon studied is: "the patient improves after 6 days".

Among the variables that can be incorporated into the study are age, weight, the existence of a chronic disease or not, the number of days since the patient has been ill, etc.

So without the quantitative and qualitative methods, we cannot confirm that chloroquine is an effective treatment for coronavirus.

Outcome differences between quantitative and qualitative research

An analysis is said to be qualitative when the phenomenon is measured by a category, a yes/no answer, for example: Increase or decrease in inflation, a patient has diagnosed positive or negative about a disease, etc.

Identify design from a description of processes ( I am not really sure what they are asking)

•   The quantitative and qualitative methods follow the following process: 
  
• -  Definition of temporal space and an empirical field
• -  Observation of the variable understudy 
• -  Organization and constitution of the database, samples, etc.
• -  Pre-processing, data cleansing, processing of missing data
• -  Modeling, interpretation, validation
• -  Reporting.

Errors in data entry that you find on statistical results output—what do you do?

Errors in the statistical results generally occur when the data pre-processing step has not been properly done.

Among other things, an aberration due to the presence of missing data may be found, in which case either remove it or replace it with another value (usually the mean).

As we can find outliers in the data (false data, a value very different from the others,...) in this case we have to replace them because they distort the expected results.

Reporting results of statistical analysis-what are critical factors?

The importance of not over-interpreting data is a key factor

Matching to research question and analysis

Types of rigor-reliability and validity and how they are used.

To ensure that quantitative/qualitative modeling meets our concerns, the use of test data is generally used.

Test data are part of the observations collected but have not been incorporated into the process, but have been set aside to serve as a means of validating the veracity of the results obtained.

This is used specially when the results are extended for the purpose of predicting new observations in the future.

Among the measures used is the false-positive rate, which is the presence of cases that are actually positive but classified by the model as negative. This measure should be minimal.

True-negative rate, which is the rate of cases that are actually negative but the statistical model classifies them as positive. This measure must be minimal.

Levels of Data-definitions and characteristics of each level

How can varying variables be measured at different levels and the manner in which it measures dictates the level of those data collected?

This is a qualitative variable coded in three categories but with a quantitative aspect related to the fact that the coding increases with the increase in temperature.