Introduction

In the blog, https://researchmethodsclass.com/research-methodology/, we said that research methodology encompasses the following:

1. Research Approach – Quantitative, Qualitative, Multi or Mixed???
2. Research Designs – plan to address the approach
3. Population, Sample Size and Sampling Procedures/Techniques
4. Methods of Data Collection – Development, administration in terms of when and how data will be collected (Data Collection Procedure), piloted, determining validity and reliability
5. Methods of Data Analysis – Statistical, Inductive or both????

Before beginning a study, a researcher must decide how the research problem will be investigated. In other words, the researcher must lay a detailed plan of how the study will answer the research problem. This broad plan is known as the research approach. The approach influences the kinds of research questions asked, the type of research questions asked, the type of data collected, the methods of analysis used, and the way conclusions are developed.

Let’s first distinguish between an approach, research method, design, and tools. These terms are related, but they are not interchangeable.  

• An approach is the strategic direction of the research. It is the broad orientation guiding the study. There are various categories of approaches in research. For instance:
• The approach based on the extent to which the findings apply to social science settings, i.e., the purpose or utility of research (e.g., basic vs. applied research);
• The approach based on the methods used to design the study and to collect data (e.g., qualitative vs. quantitative approaches). A design is the overall plan for answering the research questions.
• Methods are procedures for collecting or analyzing data,
• Tools are the instruments used to implement those methods — for example, a questionnaire, interview guide, observation checklist, or achievement test

This is the most common research approaches are quantitative research, qualitative research and mixed methods research. Quantitative research collects numerical data; qualitative research explores meanings, experiences, and processes; and mixed methods research intentionally combines quantitative and qualitative evidence in one study. No approach is automatically superior to the others. The most appropriate choice depends on the research problem, objectives, questions, philosophical assumptions, available resources, ethical considerations, and the type of evidence needed.

This blog will focus on a detailed explanation of the quantitative research approach.

Quantitative Research Approach

Quantitative research is the systematic collection, analysis & interpretation of numerical data in order to produce results that can be generalized to the target population. This approach investigates phenomena through the systematic collection and analysis of numerical data; i.e., it deals with variables that can be expressed in terms of quantity

Quantitative research most often uses deductive reasoning or logic, in which researchers start with hypotheses and then collect data which can be used to determine whether empirical evidence to support that hypothesis exists. Quantitative research measures variables, describes patterns, examines relationships, compares groups, tests hypotheses and, where the design permits, estimates the effects of an intervention. Data may be collected using structured questionnaires, tests, experiments, Likert scales, rating scales or structured observations.

Key characteristics

1. Measurement of variables: measures variables that can be expressed in terms of quantity. Concepts such as achievement, attitudes, participation, or satisfaction that are not tangible are operationalised into measurable indicators and, in most cases, measured using scales like the Likert scale, motivation scale, etc
2. Numerical data: Information is represented as counts, scores, percentages, rates, measurements, or other quantities.
3. Structured procedures: Data collection often follows standardized instruments and consistent instructions so that participants are measured in comparable ways. The instruments are prepared and pilot tested before going to the field. Ideally, the researcher should not make adjustments in the field
4. Deductive reasoning: Many quantitative studies begin with a theory or hypothesis and examine whether the collected evidence is consistent with expectations.
5. Statistical analysis: Analysis of numerical data is done through the use of statistics. Descriptive statistics summarize data, while inferential statistics are used to estimate population characteristics or test hypotheses. The use of inferential statistics allows the researcher to generalize the findings to the wider population.
6. Sampling and generalization: Probability sampling is used in quantitative research, and the sample is representative of the population to allow generalization.

Common quantitative research designs

1. Survey

To survey is to question people and record their responses for analysis.

A survey is a non-experimental research design used to describe the characteristics of a population by use of a survey or questionnaire. That is why some researchers talk about descriptive design.

NB: Descriptive is a type of data analysis

A survey is an attempt to collect data from members of a population in order to determine the current status of the population with respect to one or more variables. (Christopher, Harriet & Peter 2016).  Surveys include questions in which research participants report about their attitudes, preferences, opinions, beliefs, activities, emotions, thoughts, and behaviors in a systematic manner. It documents the characteristics, distribution or current status of a population or phenomenon.

It may answer questions such as what, how many, how often, or what proportion.

Types of Survey Designs

a) Correlational Survey Design

A correlational design examines the direction and strength of relationships between two or more variables without manipulating them. A study of the relationship between lecturer feedback frequency and student satisfaction is an example. Correlation may be positive, negative, or absent, but it should not be interpreted as proof of cause and effect.

• Cross-Sectional Survey Designs

A cross-sectional study collects data at one point or within a short period and provides a snapshot of a phenomenon. Such surveys give an idea about how things are for the respondents at a particular point in time when the survey is administered.

• Longitudinal Survey Design

A longitudinal study is a follow-up study that collects data repeatedly over time to examine change, development, or trends. Longitudinal studies can provide evidence about temporal patterns but may face attrition, cost, and complex data-management challenges.

2. Causal-Comparative (Ex Post Facto) Design

Ex Post implies ‘from after the fact ’. “After the fact: means the changes in the dependent variable are observable. The main concern of the researcher is to determine the factors that brought about the changes; i.e., the researcher investigates a problem by studying the variables in retrospect. It is interested in determining whether there is a relationship between variables that is causal in nature (i.e., does one variable cause a given variable).

This design is adopted when many of the cause-and-effect relationships under study do not permit experimental manipulation. WHY? Because the researcher DOES NOT AND SHOULD NOT manipulate the IV. The researcher deals with already formed groups. Thus, the differences between the groups are not brought on by the researcher.  

For example, a researcher may compare achievement among students who have access to digital devices and others who do not have access to digital devices. Because the researcher does not assign participants to the conditions, conclusions about causation (i.e., access to digital devices causes higher/lower achievement among students) must be made cautiously.

3. Experimental Design

Experimental studies attempt to establish a cause-and –effect relationship. A cause-and –effect relationship refers to intervention A causing behavior B to occur. Therefore, experimental designs allow the researcher to determine the effects of a given intervention on a variable in a setting that is controlled by the researcher.

An experimental design involves deliberate manipulation of an independent variable and uses random selection and random assignment to assign individuals to the experimental and the control group. The effects of an intervention are ascertained by comparing the behavior of the group that is participating in the intervention (the treatment or experiment group) with that of the selected group that is not participating in the intervention (the control group).

For instance, a researcher might randomly select and assign eligible students to an intervention group with access to digital devices and a comparison group with no digital devices to measure learner achievement. After some time, time, the comparison of the two would determine whether access to digital devices cause an increase/decrease in learner achievement.

HOWEVER:  Causality is established ONLY WHEN the researcher has controlled for EXTRANEOUS VARIABLES

There are two types of Experimental Designs:

1. True Experimental
2. Quasi Experimental

The next upcoming blogs will focus on types of experimental design and how to control for extraneous variables.