Unit 2: Research Process and Research Design - Subjective Questions

The research process is a systematic series of activities undertaken to identify a problem, collect relevant information, analyze evidence, and draw valid conclusions.

The major steps are:

• Identification of the research problem: Select and clearly define the issue to be studied.
• Review of literature: Examine books, journals, reports, and previous studies related to the problem.
• Formulation of objectives and hypotheses: State what the study intends to achieve and develop testable assumptions where appropriate.
• Selection of research design: Choose a suitable plan for collecting and analyzing data.
• Sampling: Select a representative group from the target population.
• Data collection: Gather primary or secondary data using appropriate tools.
• Data processing and analysis: Edit, code, classify, tabulate, and analyze the collected data.
• Interpretation of findings: Relate the results to the objectives and research questions.
• Report writing and presentation: Present the methodology, findings, conclusions, and recommendations in an organized report.

These steps are interrelated, and researchers may return to earlier steps when new information requires modification of the study.

Identifying and formulating a research problem is the foundation of a successful study. A clearly defined problem gives direction to every subsequent research activity.

Its importance includes:

• Provides direction: It clarifies what the researcher will investigate.
• Defines the scope: It prevents the study from becoming too broad or unfocused.
• Guides the objectives: Research objectives are developed from the identified problem.
• Determines the methodology: The nature of the problem influences the research design, sampling method, and data collection tools.
• Supports feasibility assessment: The researcher can determine whether sufficient time, money, participants, and information are available.
• Avoids duplication: Reviewing existing studies helps identify gaps and prevents unnecessary repetition.
• Improves relevance: A well-formulated problem ensures that the study addresses a meaningful academic or practical issue.

A good research problem should be clear, specific, researchable, significant, and ethically acceptable.

A review of literature is a systematic examination and analysis of existing books, scholarly articles, reports, theories, and research findings related to a particular topic.

Its major roles are:

• Provides background knowledge: It helps the researcher understand the concepts and theories associated with the topic.
• Identifies research gaps: It reveals questions that previous researchers have not adequately addressed.
• Prevents duplication: It shows what has already been studied.
• Develops a theoretical framework: Existing theories can be used to explain relationships among variables.
• Refines the research problem: The original topic can be narrowed and made more precise.
• Assists in method selection: Previous studies provide information about suitable research designs and instruments.
• Supports interpretation: Findings can be compared with results from earlier studies.
• Establishes the significance of the study: The review demonstrates why new research is needed.

A good literature review should be critical and analytical rather than merely a list of summaries.

Research objectives, research questions, and research hypotheses are related but serve different purposes.

• Research objectives: These state the specific outcomes that the study intends to achieve. They usually begin with verbs such as to examine, to compare, to determine, or to assess.
• Research questions: These are specific questions that the study seeks to answer. They guide data collection and analysis, especially in exploratory and qualitative studies.
• Research hypotheses: These are tentative, testable statements predicting a relationship or difference between variables. They are commonly used in quantitative and experimental studies.

For example, a study may have the following elements:

• Objective: To examine the effect of online learning on academic performance.
• Research question: How does online learning affect academic performance?
• Hypothesis: Students who use online learning resources regularly achieve higher academic performance than students who do not.

Thus, objectives describe the intended purpose, questions identify what must be answered, and hypotheses make predictions that can be empirically tested.

Primary data are collected directly by the researcher for the specific purpose of the current study. Examples include questionnaire responses, interviews, observations, and experiments.

Advantages of primary data:

• They are directly relevant to the research objectives.
• The researcher controls the method and quality of collection.
• Current and specific information can be obtained.

Limitations of primary data:

• Collection may be time-consuming and expensive.
• It may require trained investigators.
• Non-response and researcher bias can affect quality.

Secondary data are data previously collected by another person or organization for a different purpose. Examples include census reports, government publications, company records, books, and journal articles.

Advantages of secondary data:

• They are usually less expensive and faster to obtain.
• They help provide background information.
• They may cover large populations or long time periods.

Limitations of secondary data:

• They may not exactly match the current research objectives.
• Their accuracy and reliability may be difficult to verify.
• They may be outdated or incomplete.

Researchers should assess the relevance, reliability, adequacy, and timeliness of secondary data before using them.

A research design is the overall plan, structure, and strategy used to conduct a research study. It specifies how data will be collected, from whom, when, and how the data will be analyzed to answer the research questions.

Research design is important because it:

• Provides a clear plan: It organizes the major activities of the investigation.
• Connects objectives and methods: It ensures that the chosen methods are appropriate for the research problem.
• Controls bias: A well-planned design reduces systematic errors and threats to validity.
• Improves reliability: Standardized procedures make results more consistent.
• Guides sampling and data collection: It helps determine the population, sample, instruments, and procedures.
• Saves resources: Careful planning prevents unnecessary collection of irrelevant data.
• Supports valid conclusions: It helps the researcher distinguish genuine findings from results caused by confounding factors.
• Facilitates replication: Other researchers can understand and repeat the study.

The design should be selected according to the purpose of the study, the nature of the problem, available resources, and ethical requirements.

A good research design provides a reliable and efficient framework for answering research questions. Its main characteristics include:

• Clarity: The design clearly states the objectives, variables, population, procedures, and methods of analysis.
• Suitability: It is appropriate for the nature and purpose of the research problem.
• Validity: It enables the researcher to measure what the study intends to measure.
• Reliability: It produces consistent results when similar procedures are repeated.
• Objectivity: It minimizes personal bias in data collection and interpretation.
• Control: It manages extraneous variables and alternative explanations where necessary.
• Flexibility: It can accommodate reasonable changes when new information emerges, particularly in exploratory research.
• Economy: It uses time, money, and human resources efficiently.
• Feasibility: The design can realistically be implemented with the available resources.
• Ethical acceptability: It protects participants and respects informed consent, privacy, and confidentiality.

A strong design balances scientific rigor with practical and ethical considerations.

Research designs can be classified according to the purpose of the investigation.

• Exploratory research design: This is used when the problem is not clearly understood. It helps generate ideas, identify variables, and develop research questions. Common methods include literature review, expert interviews, focus groups, and case studies.
• Descriptive research design: This describes the characteristics of a population, event, or situation. It answers questions such as who, what, when, where, and how much. Surveys, observation, and descriptive case studies are common methods.
• Explanatory research design: This investigates why something occurs and examines relationships between variables. It attempts to explain cause-and-effect relationships using methods such as experiments, correlational studies, and statistical analysis.

The designs differ in purpose:

• Exploratory research seeks initial understanding.
• Descriptive research provides an accurate account of existing conditions.
• Explanatory research provides reasons and causal explanations.

A single research project may use more than one design in sequence.

An experimental research design is a design in which the researcher deliberately manipulates one or more independent variables to observe their effect on a dependent variable while controlling other factors.

Its essential elements are:

• Independent variable: The factor manipulated by the researcher.
• Dependent variable: The outcome measured to determine the effect of the manipulation.
• Experimental group: Participants who receive the treatment or intervention.
• Control group: Participants who do not receive the treatment or receive a standard condition.
• Random assignment: Participants are assigned to groups by chance to reduce selection bias.
• Control of extraneous variables: Other influences are minimized or held constant.
• Measurement: Outcomes are measured using valid and reliable instruments.

Experimental designs are useful for testing causal relationships. For example, a researcher may examine whether a new teaching method improves examination performance. A major strength is strong control over variables, while limitations may include ethical restrictions, artificial settings, and difficulty generalizing results to real-life situations.

Qualitative and quantitative research designs differ in their approach to investigating a problem.

Qualitative research is suitable when the researcher wants to understand how and why people experience a phenomenon. Quantitative research is suitable when the researcher wants to measure how much, identify patterns, or test hypotheses. Mixed-method research combines both approaches to obtain complementary evidence.

A cross-sectional design collects data from participants at one point in time or during a short period. It provides a snapshot of characteristics, attitudes, or conditions.

Advantages of cross-sectional design:

• Relatively quick and economical.
• Useful for estimating prevalence and describing current conditions.
• Can compare different groups at the same time.

Limitations of cross-sectional design:

• Cannot clearly establish changes over time.
• Provides limited evidence about causal relationships.
• Age or group differences may be confused with time-related changes.

A longitudinal design collects data from the same participants or population repeatedly over an extended period.

Advantages of longitudinal design:

• Measures development and change over time.
• Helps establish temporal order between variables.
• Provides stronger evidence about long-term effects.

Limitations of longitudinal design:

• Requires more time and financial resources.
• Participants may drop out, creating attrition bias.
• Repeated measurement may influence participants' responses.

The choice depends on whether the study requires a current description or an analysis of change and development.

A mixed-methods research design combines qualitative and quantitative approaches within one study. The two types of data may be collected sequentially or simultaneously and are integrated during interpretation.

Common forms include:

• Convergent design: Qualitative and quantitative data are collected separately and then compared or combined.
• Explanatory sequential design: Quantitative data are collected first, followed by qualitative data to explain unexpected or detailed findings.
• Exploratory sequential design: Qualitative exploration is conducted first and is followed by quantitative testing or measurement.

Mixed-methods research is appropriate when:

• A problem requires both numerical measurement and contextual understanding.
• Quantitative findings need explanation through participant experiences.
• Qualitative themes need to be tested with a larger sample.
• Different forms of evidence can strengthen confidence in the conclusions.

Its main advantage is complementarity, because one method can address limitations of the other. However, it requires more time, expertise, resources, and careful integration of findings.

Selecting a research design involves matching the study plan with the research problem and objectives. The major stages are:

• Clarify the research problem: Determine exactly what issue is being investigated.
• Specify the purpose: Decide whether the study is exploratory, descriptive, explanatory, evaluative, or experimental.
• Formulate research questions or hypotheses: Identify the information that must be obtained and tested.
• Identify the variables and concepts: Define the independent, dependent, and control variables where applicable.
• Select the research approach: Choose qualitative, quantitative, or mixed methods.
• Determine the time dimension: Decide whether the study will be cross-sectional or longitudinal.
• Choose the level of control: Assess whether an experiment, survey, case study, or observational design is most suitable.
• Consider the population and sample: Ensure that the design can reach the required participants.
• Assess feasibility and ethics: Consider cost, time, access, risks, consent, privacy, and confidentiality.
• Plan data analysis: Select procedures that can answer the research questions.

The final design should be scientifically appropriate, practically feasible, and ethically defensible.

The terms population and sample describe the groups to which a research study relates.

• Population: The complete set of individuals, objects, events, or records possessing the characteristics relevant to a study.
• Target population: The entire group to which the researcher intends to generalize the findings. For example, all undergraduate students in a country may be the target population.
• Accessible population: The portion of the target population that the researcher can realistically reach because of geographic, administrative, or time-related limitations.
• Sample: A selected subset of the population from which data are actually collected.

The relationship can be represented as:

Target population -> Accessible population -> Sample

A study should define these groups clearly because the quality of generalization depends on how closely the sample represents the target population. Confusing the target population with the accessible population can lead to inaccurate conclusions about whom the findings apply to.

Sampling is the process of selecting a subset of units from a population for the purpose of collecting data and drawing conclusions about the larger population.

Researchers use samples because:

• Census may be impractical: Studying every population member may be impossible when the population is large or widely dispersed.
• It saves time: Data can be collected and analyzed more quickly.
• It reduces cost: Fewer participants require fewer financial and human resources.
• It allows detailed investigation: Researchers can devote more attention to each selected unit.
• Some studies are destructive: Testing may damage or consume the unit being studied.
• Results may still be reliable: A properly selected representative sample can provide accurate estimates of population characteristics.

However, sampling introduces the possibility of sampling error and bias. To reduce these problems, researchers must define the population clearly, use an appropriate sampling frame, select a suitable sampling technique, and ensure an adequate sample size.

The main distinction is whether every population member has a known and non-zero chance of being selected.

Probability sampling:

• Selection is based on random procedures.
• Each unit has a known or calculable chance of selection.
• Sampling error can be estimated statistically.
• Findings can usually be generalized more confidently to the population.
• Examples include simple random, systematic, stratified, and cluster sampling.

Non-probability sampling:

• Selection is based on convenience, judgment, quotas, referrals, or researcher access.
• The probability of selection is unknown for at least some population members.
• Sampling error cannot be estimated reliably using standard probability formulas.
• Generalization is limited, although the method may be useful for exploratory or hard-to-reach populations.
• Examples include convenience, purposive, quota, and snowball sampling.

Probability sampling is generally preferred when representativeness and statistical generalization are important. Non-probability sampling may be preferred when time, access, or the nature of the population makes random selection difficult.

The major probability sampling techniques are as follows:

• Simple random sampling: Every population member has an equal chance of selection. Names may be chosen using a random number generator or a lottery method.
• Systematic sampling: The researcher selects every th unit from an ordered sampling frame after choosing a random starting point. The sampling interval is commonly calculated as , where is the population size and is the required sample size.
• Stratified sampling: The population is divided into homogeneous subgroups called strata, such as gender, age, or region. Samples are then selected from each stratum, often in proportion to its population size.
• Cluster sampling: The population is divided into naturally occurring groups called clusters, such as schools, villages, or districts. A number of clusters are randomly selected, and all or some units within them are studied.

Simple random sampling is conceptually straightforward, stratified sampling improves subgroup representation, systematic sampling is operationally convenient, and cluster sampling is useful when the population is geographically dispersed.

These are common non-probability sampling techniques.

• Convenience sampling: Participants are selected because they are easily available. It may be used for pilot studies or classroom research, but it can produce substantial bias.
• Purposive sampling: The researcher deliberately selects participants who possess specific knowledge, characteristics, or experiences. It is useful in qualitative studies and expert research.
• Quota sampling: The population is divided into categories, and participants are selected conveniently until a required number, or quota, is obtained from each category. It can improve subgroup balance but does not provide random selection.
• Snowball sampling: Existing participants refer the researcher to other eligible participants. It is useful for hidden, rare, or difficult-to-reach populations, such as undocumented workers or members of stigmatized groups.

These methods are useful when a complete sampling frame is unavailable or when the study requires specialized participants. Their major limitation is restricted representativeness and limited ability to generalize findings statistically.

A sampling frame is a complete operational list or source that identifies the members of the population from which a sample is selected. Examples include an employee register, electoral roll, school enrollment list, or customer database.

An adequate sampling frame should have the following qualities:

• Completeness: It should include all, or nearly all, eligible population units.
• Accuracy: Names, addresses, and other identifying information should be correct.
• Currency: It should be updated to reflect additions, removals, and changes in membership.
• No duplication: Each population unit should appear only once.
• Clear eligibility: It should be possible to determine which units belong to the target population.
• Accessibility: The researcher should be able to use it lawfully and practically.
• Appropriate coverage: It should correspond closely to the target population rather than to a different group.

An incomplete frame causes undercoverage, while duplicate entries can give some units a greater chance of selection. Both problems may reduce the representativeness of the sample.

Sampling error is the difference between a characteristic calculated from a sample and the true characteristic of the population. It occurs because only a subset of the population is studied. Sampling error can be reduced by:

• Increasing the sample size where appropriate.
• Using probability sampling.
• Ensuring that the sampling frame adequately covers the population.
• Using stratification when important subgroups differ substantially.

Non-sampling error arises from problems unrelated to selecting a sample. Examples include:

• Coverage error: Some population members are omitted from the sampling frame.
• Non-response error: Selected participants do not provide data.
• Measurement error: Questions or instruments do not measure the intended concept accurately.
• Interviewer error: The investigator influences responses or records data incorrectly.
• Processing error: Data are incorrectly coded, entered, or analyzed.

Non-sampling errors can be reduced through careful instrument design, pilot testing, interviewer training, follow-up contacts, quality checks, clear procedures, and appropriate data validation.