Free preview

Introduction to nursing research

DNE 122 · Applied Research and Teaching Methodology — Diploma Nursing - Extension

Learning objectives

  • Introduction to research
  • Terminologies
  • Research Ethics
  • Purpose of studying research
  • Research techniques (Qualitative, quantitative and their approaches)
  • Research designs

Introductory study notes

Introduction to research

Example in nursing

A nurse notices that some patients fail to complete their prescribed antibiotics after discharge from hospital.

The nurse may conduct research to investigate:

  • Why patients do not complete their medication.

  • Whether patients understand their medication instructions.

  • The influence of medication costs and availability.

  • The relationship between health education and medication adherence.

The findings may help healthcare workers develop interventions to improve treatment outcomes.

Breakdown of the concept

Component

Meaning

Systematic

Follows an organized sequence of activities.

Scientific

Uses logical, evidence-based methods.

Investigation

Examines a question or problem.

Data collection

Obtains information from appropriate sources.

Analysis

Examines and organizes collected information.

Interpretation

Explains the meaning and implications of findings.

Knowledge generation

Produces new information or validates existing knowledge.

Definitions of research

Research can be understood through several perspectives:

A. Research as a systematic investigation

Research is a planned investigation conducted to establish facts, relationships, or explanations.

B. Research as a scientific process

Research uses scientific methods to develop evidence that can support or challenge existing knowledge.

C. Research as problem-solving

Research identifies problems and investigates possible solutions using evidence.

D. Research as knowledge development

Research contributes to the development of theories, practices, policies, and interventions.

Important examination point: Research is not simply searching for information on the internet. It involves a structured process for answering a defined question using appropriate methods.

The meaning of scientific research

Scientific research is a disciplined approach to investigating phenomena through systematic observation, measurement, analysis, and logical reasoning.

It may be conducted in a laboratory, hospital, community, classroom, or other setting.

Main features of scientific research

  1. Empirical: Based on observations or evidence.

  2. Systematic: Follows a planned procedure.

  3. Logical: Uses rational reasoning.

  4. Objective: Attempts to minimize personal bias.

  5. Critical: Examines evidence and alternative explanations.

  6. Verifiable: Methods and findings can be examined or evaluated by others.

  7. Ethical: Protects participants and promotes responsible research conduct.

Characteristics of research

1. Systematic

Research follows an organized process, from identifying a problem to communicating findings.

Example: A nurse develops a research protocol before collecting data about patient satisfaction.

2. Scientific

Research uses established scientific methods and appropriate evidence.

Example: A researcher uses a validated questionnaire to measure knowledge of HIV prevention.

3. Empirical

Research is based on information obtained through observation, measurement, interviews, experiments, or other evidence-gathering methods.

4. Controlled where appropriate

Researchers attempt to manage factors that may affect findings. The extent of control depends on the research design.

For example, an experimental study may compare an intervention group with a control group, whereas an ethnographic study may focus on observing people's experiences in their natural environment.

5. Objective

Researchers should strive to reduce bias and report findings honestly, including results that do not support their expectations.

6. Replicable or reproducible where appropriate

Research methods should be described sufficiently for other researchers to evaluate or repeat the investigation, although exact replication is not possible or expected in every qualitative or contextual study.

7. Problem-oriented

Research generally begins with a question, knowledge gap, or problem requiring investigation.

8. Analytical

Researchers examine relationships, patterns, themes, or differences within collected information.

9. Ethical

Research must respect human dignity, safety, rights, privacy, and applicable regulations.

10. Communicable

Research findings should be presented clearly so that other people can understand and potentially apply the knowledge.

The research process

The research process is the sequence of activities undertaken to conduct a research study.

Research process flowchart

1. Identify the research problem

2. Review the literature

3. Formulate objectives and research questions

4. Select research design and methods

5. Obtain ethical approval and permissions

6. Collect data

7. Analyze data

8. Interpret findings

9. Draw conclusions and recommendations

10. Disseminate and apply findings

Note: Research is often iterative. Researchers may revise questions, methods, or interpretations as knowledge develops.

Step 1: Identification of the research problem

A research problem is a situation, question, or knowledge gap that requires systematic investigation.

Sources of research problems:

  • Clinical practice.

  • Personal observations.

  • Existing literature.

  • Community health challenges.

  • Previous research recommendations.

  • Healthcare policies.

  • Theoretical frameworks.

  • Institutional reports.

Example: Low attendance at antenatal care clinics in a particular community.

A researcher may investigate the factors contributing to low attendance.

Step 2: Review of literature

A literature review involves searching for, evaluating, and synthesizing existing knowledge related to the research problem.

It helps the researcher understand what is already known and what remains uncertain.

Purposes:

  1. Identify knowledge gaps.

  2. Avoid unnecessary duplication.

  3. Develop the research framework.

  4. Identify suitable research methods.

  5. Support the rationale for the study.

  6. Compare findings with previous studies.

Step 3: Formulation of objectives

Research objectives describe what the study intends to achieve.

Example:

General objective: To assess factors associated with medication non-adherence among patients with hypertension attending a health facility.

Specific objectives:

  1. To determine the level of knowledge about prescribed antihypertensive medication.

  2. To identify social and economic factors affecting adherence.

  3. To assess the association between medication knowledge and adherence.

Step 4: Selection of research design

A research design is the overall plan for conducting a study.

Examples include:

  • Descriptive cross-sectional design.

  • Case-control design.

  • Cohort design.

  • Experimental design.

  • Phenomenological design.

  • Qualitative descriptive design.

The design must match the research question, resources, ethical considerations, and intended conclusions.

Step 5: Ethical approval and permission

Before beginning research involving human participants, researchers must determine the required ethical review and administrative approvals.

Activities may include:

  • Submitting a research protocol to an appropriate research ethics committee.

  • Obtaining institutional permission.

  • Preparing informed consent documents.

  • Planning confidentiality and data protection.

  • Developing procedures for managing participant risks.

Step 6: Data collection

Data collection involves obtaining the information required to answer the research questions.

Examples:

  • Interviews.

  • Questionnaires.

  • Observation.

  • Medical record review.

  • Laboratory measurements.

  • Focus group discussions.

Step 7: Data analysis

Data analysis involves organizing and examining information to identify meaningful findings.

Quantitative data may be analyzed using frequencies, percentages, means, and statistical tests.

Qualitative data may be analyzed through coding, categorization, and thematic analysis.

Step 8: Interpretation

Interpretation explains what the findings mean in relation to the research objectives, existing evidence, and limitations.

Step 9: Conclusion and recommendations

The conclusion summarizes the findings and answers the research questions.

Recommendations should be based on the evidence and relevant to practice, policy, education, or future research.

Step 10: Dissemination

Dissemination means communicating research findings to relevant audiences.

Methods include:

  • Research reports.

  • Academic presentations.

  • Scientific publications.

  • Posters.

  • Workshops.

  • Policy briefs.

  • Clinical guidelines.

Types of research

Research can be classified according to its purpose, approach, design, and application.

A. Classification according to approach

Research types

1. Quantitative research

Uses numerical data and statistical methods to measure variables, examine patterns, or test hypotheses.

Example: Measuring the proportion of nurses who follow hand hygiene guidelines.

Quantitative research

2. Qualitative research

Explores experiences, perceptions, beliefs, behaviors, and meanings using non-numerical data.

Example: Exploring the experiences of mothers caring for children with sickle cell disease.

Qualitative research

3. Mixed-methods research

Combines quantitative and qualitative approaches in one research project to address complementary aspects of a question.

Example: Measuring missed clinic appointments and interviewing patients about why appointments are missed.

B. Classification according to purpose

Type

Explanation

Example

Basic research

Develops theoretical knowledge or understanding.

Investigating mechanisms of inflammation.

Applied research

Addresses a practical problem.

Evaluating a strategy to improve medication adherence.

Action research

Systematically investigates a problem while implementing and evaluating changes in practice.

Nurses introduce a new patient education process and evaluate it.

Evaluation research

Assesses the implementation, quality, effectiveness, or outcomes of a program or intervention.

Evaluating a maternal health education program.

Exploratory research

Investigates a topic where understanding is limited.

Exploring nurses' experiences with a new electronic health record.

Descriptive research

Describes characteristics, frequencies, or patterns.

Describing the distribution of SCD admissions in a hospital.

C. Classification according to time

Cross-sectional research

  • Collects information at one point in time or within a defined short period.

  • Commonly used to estimate prevalence and describe characteristics.

  • Does not automatically establish causation.

Longitudinal research

  • Collects data from participants or populations over an extended period.

  • Can examine changes and temporal relationships.

Examples of longitudinal designs include:

  • Cohort studies.

  • Panel studies.

  • Repeated-measures studies.

Retrospective research

- Examines information about past events using existing records or participants' recall.

Prospective research

- Follows participants or collects information forward from the study's starting point.

More types of research (frequently examined)

The classifications above overlap, and different textbooks use different names for the same type of research. The types below are commonly examined and should be known in addition to the ones already listed.

Basic (pure or fundamental) research

Basic research is done mainly to expand knowledge and understanding, not to solve an immediate practical problem. It lays the foundation for applied research and usually has no immediate commercial or practical use.

Applied research

Applied research is designed to solve practical, present-day problems and provide information that can be used immediately by programmes, hospitals or communities. Evaluation research, operational research and action research are all forms of applied research.

Evaluation research

Evaluation research assesses whether a programme, policy or legal change has achieved its intended outcomes and what impact it has had. Its results are used by organisations in the near future to continue, change or stop a programme.

Operational (operations) research

Operational research is "research on operations". It applies a scientific, usually quantitative and mathematical, approach to problems of how activities in an organisation should be run and coordinated, so that scarce resources are used effectively.

  • It began during World War II, when British and American military managers asked scientists to help allocate limited resources between military operations.

  • After the war its success led to its use in industry, health services and government.

  • Recommendations are usually derived from mathematical models; computer software has made it easier today.

Action research (participatory action research)

Action research aims to bring about change. The people directly affected by the problem take an active part in the whole research process, from design to presenting results, so that they are empowered to make decisions and take action. It has a social purpose: empowering people of low status in an organisation or community.

Historical research

Historical research establishes facts about past events in order to understand the present more accurately. It relies on records, documents and archives.

Strengths:

  • It is unobtrusive: the research does not affect what is being studied.

  • It is well suited to analysing trends over time.

  • It provides background context for a current problem.

  • There is usually no researcher–subject interaction that could bias findings.

  • Historical sources can be reused to study different problems or to replicate a study.

Limitations:

  • Findings depend on the amount and quality of records available; gaps in documentation must be acknowledged.

  • Past data cannot be manipulated or controlled, so internal validity is weak.

  • Interpreting historical sources is time consuming.

  • Sources must be properly archived to remain accessible, which can be a problem for digital sources.

  • Original authors' biases are hard to detect.

Descriptive, exploratory and explanatory research compared

Feature

Descriptive

Exploratory

Explanatory

Purpose

Gives an accurate picture of what exists

Becomes familiar with a new or little-studied area

Explains why things are the way they are

Questions answered

Who, what, when, where, how many

What is happening? How?

Why?

Prior knowledge

Researcher already has a general understanding

Little is known

Enough is known to test explanations

Approach

Mostly quantitative (surveys, census)

Inductive, mostly qualitative, few cases

Deductive (general to specific), mostly quantitative, many cases

Output

Prevalence, distributions, characteristics

Ideas, themes, hypotheses for later studies

Causes and reasons, tested hypotheses

Advantages of exploratory research:

  • Gives background information on a new topic.

  • Is flexible and can address "what" and "how" questions.

  • Helps define new terms and clarify concepts.

  • Generates hypotheses and more precise research problems.

  • Helps policy makers set research priorities.

Limitations of exploratory research:

  • Small samples mean findings cannot usually be generalised.

  • It provides insight but not definitive conclusions.

  • Its flexible, often unstructured process gives tentative results of limited value to decision-makers.

Health sciences (health systems) research

Health sciences research focuses on health and medical problems. Its characteristics are:

  • A focus on real problems in health and health services.

  • Practical, timely results that can be used by decision-makers.

  • A multidisciplinary approach, involving health workers, social scientists, statisticians and managers.

Research designs

A research design is the framework that determines how a study will be organized, how data will be collected, and how research questions will be answered.

1. Descriptive research design

Describes characteristics of a population, condition, or situation.

Example: A study describing the prevalence of anemia among pregnant women attending a health center.

Advantages:

  • Useful for identifying health problems.

  • Can provide baseline information.

  • Often practical and relatively affordable.

Limitations:

  • May not explain why a problem occurs.

  • Cross-sectional descriptive studies generally cannot establish temporal causation.

2. Correlational research design

Examines the relationship between two or more variables without necessarily manipulating them.

Example: Investigating the relationship between workload and job satisfaction among nurses.

Important: A correlation does not prove that one variable causes another.

3. Experimental research design

Researchers deliberately introduce an intervention and compare outcomes, often using control or comparison groups.

Example: Testing whether a structured health education program improves diabetes self-management.

Key concepts:

  • Intervention.

  • Control or comparison group.

  • Outcome measurement.

  • Randomization, where applicable.

  • Allocation concealment and blinding, where appropriate.

4. Quasi-experimental design

Evaluates an intervention but lacks one or more features of a true randomized experiment, such as random allocation.

Example: Comparing patient waiting times before and after implementation of a triage protocol.

5. Case-control design

Compares participants with an outcome or condition (cases) with those without it (controls) to investigate previous exposures.

Example: Comparing prior exposure to a risk factor among individuals with and without a particular disease.

6. Cohort design

Follows groups based on exposure status or other characteristics to assess the occurrence of outcomes over time.

Example: Following smokers and non-smokers to examine the incidence of chronic respiratory disease.

7. Phenomenological design

A qualitative approach that explores how individuals experience and make sense of a particular phenomenon.

Example: Exploring the lived experiences of patients receiving palliative care.

8. Ethnographic design

Studies the culture, practices, beliefs, and behaviors of a group or community, usually through prolonged engagement and observation.

Example: Exploring cultural beliefs influencing childbirth practices in a community.

Sources of research problems in nursing

Nurses can identify research problems from various areas of professional practice.

A. Clinical practice

Problems encountered while caring for patients.

Examples:

  • Frequent medication errors.

  • Poor wound healing.

  • Delayed recognition of patient deterioration.

  • Low adherence to infection prevention practices.

B. Nursing education

Examples:

  • Challenges faced by students during clinical placement.

  • Effectiveness of simulation-based teaching.

  • Factors influencing academic performance.

C. Nursing administration

Examples:

  • Staff shortages.

  • Workload and burnout-related concerns.

  • Patient waiting times.

  • Quality improvement.

D. Community health

Examples:

  • Immunization coverage.

  • Malnutrition.

  • Access to antenatal care.

  • Health-seeking behavior.

E. Public health and epidemiology

Examples:

  • Disease prevalence.

  • Risk factors.

  • Disease surveillance.

  • Health service utilization.

Research problem, research topic, and research title

These terms are related but different.

Term

Meaning

Example

Research topic

Broad area of interest.

Hand hygiene among nurses.

Research problem

Specific issue requiring investigation.

Inadequate adherence to hand hygiene procedures.

Research title

Specific statement identifying the study.

Factors associated with hand hygiene compliance among nurses at a selected hospital.

Research question

Question the study intends to answer.

What factors are associated with hand hygiene compliance?

Research objective

What the study intends to achieve.

To assess factors associated with hand hygiene compliance.

Terminologies

Essential research terms

1. Research

A systematic investigation designed to generate or validate knowledge and answer a defined question.

2. Researcher

An individual or team responsible for planning, conducting, analyzing, or reporting research.

Example: A registered nurse investigating barriers to hypertension treatment adherence.

3. Research participant

A person who contributes data or takes part in a research study.

The term participant is generally preferred when referring to individuals involved in human research.

4. Research population

The entire group of individuals, objects, or units that meet the study's eligibility criteria and to which the research is intended to relate.

Example: All nurses working in a selected regional referral hospital.

5. Target population

The broader population about which the researcher intends to draw conclusions.

6. Accessible population

The portion of the target population that the researcher can realistically reach and study.

7. Sample

A subset of the population selected for participation in the study.

Example: 120 nurses selected from a hospital workforce.

8. Sampling

The process of selecting a sample from a population.

9. Sampling frame

A list or other operational representation of the population from which a sample is selected.

Example: An eligible staff register used to select nurses for a study.

10. Sampling unit

The individual element or group selected at a particular stage of sampling.

11. Parameter

A numerical characteristic of a population.

Example: The actual mean age of all eligible nurses in a defined population.

12. Statistic

A numerical characteristic calculated from sample data.

Example: The mean age of 120 nurses included in a study.

More population and sampling terms

  • Study sample: the part of the accessible population that actually takes part in the study.

  • Sample size: the number of participants included in the study, for example 150 respondents.

  • Homogeneous population: members share the characteristic of interest, for example all are primigravidae aged 18–25.

  • Heterogeneous population: members differ in identifiable ways such as age, sex, religion or level of education. A heterogeneous population usually needs a larger sample or stratification.

  • Census: a study that includes every member of the population. A sample survey includes only part of the population.

  • Survey: a study that collects information about people's characteristics, opinions, attitudes, knowledge or practices, often over a wide area. It may use questionnaires, interviews, telephone, mail or electronic media and may be cross-sectional or longitudinal.

Variables

Meaning of a variable

A variable is a characteristic, attribute, or measurement that can take different values among individuals, groups, or observations.

Examples:

  • Age.

  • Sex.

  • Blood pressure.

  • Weight.

  • Knowledge score.

  • Medication adherence.

  • Length of hospital stay.

A. Independent variable

The variable considered as a possible explanatory factor, predictor, or exposure.

In experimental studies, it may be deliberately manipulated by the researcher.

Example: A structured patient education program.

B. Dependent variable

The outcome or response being measured.

Example: Medication adherence after patient education.

C. Confounding variable

A variable associated with both the exposure and outcome that can distort the observed relationship if not appropriately addressed.

Example: When investigating the relationship between exercise and blood pressure, age may confound the relationship if it is related to both exercise habits and blood pressure.

D. Extraneous variable

A variable other than the primary variables of interest that may influence the outcome.

Some extraneous variables can become confounders, depending on their relationships in the study.

E. Categorical variable

A variable that assigns observations to groups or categories.

Examples:

  • Blood group.

  • Marital status.

  • Occupation.

  • Disease status.

F. Continuous variable

A variable that can take numerical values across a range, including decimal values, within the limits of measurement.

Examples:

  • Body temperature.

  • Weight.

  • Height.

  • Serum glucose concentration.

G. Discrete variable

A variable that takes countable values, commonly whole numbers.

Examples:

  • Number of children.

  • Number of hospital admissions.

  • Number of medications prescribed.

More types of variables

Other names used in exams

  • Independent variable is also called the explanatory, predictor, causal, treatment or exposure variable. It comes first and influences the outcome.

  • Dependent variable is also called the outcome, response or criterion variable. It is the variable being predicted and usually represents the problem being studied.

Antecedent variable

A variable that occurs before the independent variable in time. It may influence the independent variable, or it may simply occur earlier without causing the outcome. Some textbooks use "antecedent" loosely as another name for the independent variable.

Intervening (mediator) variable

A variable that lies between the independent and dependent variables. It is caused by the independent variable and in turn causes the dependent variable. It explains how the cause produces the effect and is often not directly observed.

Moderator variable

A variable that changes the strength or direction of the relationship between the independent and dependent variables. The researcher selects or measures it to see whether the relationship differs across its levels.

Organismic variable

A characteristic within the participant, such as emotions, fatigue, intelligence or anxiety, that cannot be manipulated by the researcher but can affect the results.

Active and assigned (attribute) variables

  • Active variable: can be manipulated by the researcher, for example the teaching method (lecture versus demonstration) or type of dressing used.

  • Assigned (attribute) variable: cannot be manipulated because it is a characteristic of the participant, for example age, sex, parity, tribe or level of education.

Quantitative and qualitative variables

  • Quantitative variable: measured numerically, for example weight, haemoglobin level, number of antenatal visits.

  • Qualitative (categorical) variable: describes a quality or category with no numerical meaning, for example blood group, religion or marital status. Codes such as male = 1, female = 2 are labels only; the numbers cannot be added or averaged.

Confounding variable: a classic example

A study finds that bottle-fed infants have less diarrhoea than breastfed infants, which seems to suggest that bottle feeding protects against diarrhoea. However, mothers with more education are more likely to bottle-feed and also practise better hygiene. Mother's education is related to both the exposure (feeding method) and the outcome (diarrhoea). It is a confounder that has hidden, and even reversed, the true relationship.

How extraneous and confounding variables are controlled

Uncontrolled extraneous variables threaten both internal and external validity. The researcher should anticipate them and use one or more of these methods:

  • Randomisation: random selection and random allocation spread unknown extraneous factors evenly between groups.

  • Elimination (restriction): excluding the variable, for example studying only non-smokers.

  • Holding constant: keeping the variable the same for all participants, for example testing all participants at the same time of day.

  • Matching or balancing: making groups similar on the variable, for example matching cases and controls by age.

  • Statistical adjustment: controlling for the variable during analysis, for example using stratified analysis or regression.

Operational definition

An operational definition explains precisely how a concept or variable will be measured or identified in a particular study.

Example

Concept: Medication adherence.

Operational definition: The percentage of prescribed medication doses taken by a participant during a defined 30-day observation period, measured using a specified adherence assessment method.

Importance:

  • Improves clarity.

  • Promotes consistent measurement.

  • Reduces ambiguity.

  • Helps other researchers understand the methods.

  • Supports reproducibility.

Conceptual definition

A conceptual definition explains the theoretical or general meaning of a concept.

Example

Concept: Patient satisfaction.

Conceptual definition: The extent to which a patient's healthcare experience meets or aligns with their expectations and perceived needs.

Operational definition: A score obtained from a specified patient satisfaction questionnaire.

Levels (scales) of measurement

Variables are measured on one of four scales. Remember them with the word NOIR: Nominal, Ordinal, Interval, Ratio. Each level has all the properties of the level before it.

Scale

What it tells you

Examples

Best average

Arithmetic allowed

Nominal

Names or labels only; categories are mutually exclusive with no order

Sex, blood group, religion, tribe, marital status

Mode

Counting only

Ordinal

Categories in a meaningful order, but the gaps between them are unknown or unequal

Level of education, pain rated mild/moderate/severe, stage of cancer, satisfaction scale

Median (or mode)

Ranking

Interval

Order and equal, known intervals, but no true zero (zero is arbitrary)

Temperature in °C or °F, calendar year, IQ score

Mean (also median and mode); standard deviation can be calculated

Add and subtract, but not meaningful ratios

Ratio

Order, equal intervals and a true (absolute) zero

Weight, height, age, pulse rate, income, number of children

Mean

Add, subtract, multiply and divide

  • A nominal variable with only two categories, such as yes/no or male/female, is called dichotomous.

  • On an ordinal scale we know that "very happy" is more than "happy" but not by how much, so the mean is not appropriate.

  • On an interval scale 0 °C does not mean "no temperature", so 20 °C is not "twice as hot" as 10 °C.

  • On a ratio scale 0 kg means no weight, so 80 kg is twice 40 kg.

  • Nominal and ordinal data are often called categorical or qualitative. Interval and ratio data are numerical or quantitative.

Research objectives

Objectives are statements describing the intended achievements of the research study.

General objective

Describes the overall purpose of the study.

Example: To assess factors influencing adherence to prescribed antihypertensive medication among adult patients attending a selected health facility.

Specific objectives

Break the general objective into smaller, measurable tasks.

Example:

  1. To determine the proportion of patients who report adherence to prescribed antihypertensive medication.

  2. To assess patients' knowledge of their prescribed medications.

  3. To identify barriers to medication adherence.

Qualities of good objectives

A good objective should be:

  • Clear.

  • Specific.

  • Measurable where appropriate.

  • Relevant.

  • Feasible.

  • Consistent with the research problem.

Writing objectives the way examiners expect

  • The general objective (aim or purpose) states the overall intention of the study. It is usually the topic rewritten with an action verb, for example: "To establish the factors affecting the spread of bedbugs among male students in X hall, Y hospital."

  • Specific objectives break the general objective into smaller activities that together achieve it. Most proposals have two to four specific objectives (three is common).

  • Objectives must be complete, so they should state the target population and the study area.

  • Every objective must relate to the topic, and the final report must present results for every objective.

  • Use measurable action verbs: assess, determine, establish, identify, examine, evaluate, investigate, compare, describe.

  • Avoid verbs that cannot be measured: to study, to know, to understand, to learn about.

SMART objectives

  • S – Specific: addresses one issue only.

  • M – Measurable: uses verbs and terms that can be measured.

  • A – Achievable: can be done within the available time and money.

  • R – Realistic or relevant: addresses the problem at hand.

  • T – Time-bound: can be completed within a stated period.

Importance of objectives

  • They are the yardstick for measuring progress towards the intended goal.

  • They show the type of data needed and guide the choice of design, tools and analysis.

  • They give structure to the literature review, the results and the discussion.

Research questions

Research questions are specific questions that the study seeks to answer.

Example

Research topic: Hand hygiene among nurses.

Research questions:

  1. What proportion of nurses adhere to recommended hand hygiene procedures?

  2. What factors influence hand hygiene practices?

  3. What barriers do nurses report?

Types of research questions

Type

Purpose

Example

Descriptive

Describes a phenomenon.

What is the level of knowledge?

Comparative

Compares groups or conditions.

Is knowledge different between two groups?

Associational

Examines relationships.

Is workload associated with job satisfaction?

Exploratory

Investigates experiences or issues.

How do nurses experience workplace stress?

Evaluative

Assesses an intervention or program.

How does a training program affect hand hygiene knowledge?

Hypothesis

A hypothesis is a testable statement about an expected relationship, difference, or association between variables.

Example

Research question: Is patient education associated with improved medication adherence?

Research hypothesis: Patients who receive structured medication education will have higher adherence scores than patients who receive routine education.

A. Null hypothesis (H₀)

States that there is no specified relationship or difference between variables in the population, according to the hypothesis being tested.

Example: There is no difference in medication adherence between patients receiving structured education and those receiving routine education.

B. Alternative hypothesis (H₁ or Hₐ)

States that a specified relationship or difference exists.

Example: Medication adherence differs between the two groups.

C. Directional hypothesis

Predicts the direction of the relationship or difference.

Example: Structured medication education increases adherence scores.

D. Non-directional hypothesis

Predicts a relationship or difference without specifying its direction.

Example: Structured medication education is associated with medication adherence.

Note: Not all research requires a formal hypothesis. Many descriptive and qualitative studies use research questions instead.

More on hypotheses

The word comes from the Greek "hypotithenai", meaning "to put under" or "to suppose". A hypothesis is a tentative statement, not a proven truth. It is a statement, not a question, and it is mainly used in quantitative, especially experimental and analytical, studies. Descriptive studies usually do not need one.

Characteristics of a good hypothesis

  • Simple, specific, clear and stated before data collection.

  • Testable using available techniques.

  • States the expected relationship between variables, preferably in declarative form.

  • Based on a good research question and consistent with existing knowledge.

  • Has empirical referents, meaning the variables can be observed and measured.

Simple and complex hypotheses

  • Simple hypothesis: one independent and one dependent variable. Example: a sedentary lifestyle is associated with high blood pressure.

  • Complex hypothesis: more than one independent or dependent variable. Example: a sedentary lifestyle and alcohol use are associated with high blood pressure.

  • A hypothesis written as directional or non-directional is said to be in declarative form.

Testing a hypothesis

The null hypothesis assumes that any difference seen in the sample is due to chance alone. The alternative hypothesis assumes that the difference is due to a real, non-random cause. If the null hypothesis is rejected, the alternative is accepted.

Type I and Type II errors

H₀ is actually true

H₀ is actually false

H₀ rejected

Type I error (false positive): concluding there is an effect when there is none

Correct decision

H₀ not rejected

Correct decision

Type II error (false negative): missing an effect that really exists

  • The probability of a Type I error is the significance level (alpha), usually set at 0.05.

  • The risk of a Type II error is reduced by using an adequate sample size.

Reasons for using hypotheses

  • They give the study a specific focus and direct the investigation.

  • They allow the researcher to confirm or reject relationships between variables.

  • When supported, they provide evidence of predictive relationships.

  • They provide a framework for organising and summarising results and conclusions.

Literature review

A literature review is a structured process of locating, evaluating, and synthesizing relevant published and unpublished evidence.

Types of literature sources

  1. Primary research articles.

  2. Systematic reviews.

  3. Meta-analyses.

  4. Textbooks.

  5. Government publications.

  6. Clinical guidelines.

  7. Institutional reports.

  8. Theses and dissertations.

Primary and secondary sources

Primary source: Reports original research conducted by the authors.

Example: A research article presenting original data from a study of SCD patients.

Secondary source: Summarizes, evaluates, or synthesizes evidence from other sources.

Example: A systematic review of interventions for SCD pain management.

Systematic review

A review that uses predefined, transparent methods to identify, appraise, and synthesize evidence addressing a focused research question.

Meta-analysis

A statistical method that combines numerical results from multiple studies when appropriate.

A systematic review may or may not include a meta-analysis.

Research framework

Conceptual framework

A conceptual framework presents the key concepts or variables in a study and the expected relationships among them.

Example: Factors influencing medication adherence

Independent / explanatory factors

Knowledge • Medication cost • Side effects • Social support

Medication adherence

Outcome variable

Potential health outcomes

Blood pressure control • Treatment outcomes

Illustrative framework. The relationships must be specified and assessed using the chosen research design; the diagram does not establish causation.

Theoretical framework

A theoretical framework uses an established theory to explain concepts and relationships relevant to a study.

Example: The Health Belief Model may help researchers examine beliefs influencing health-seeking behavior.

Validity and reliability

Validity

Validity refers to the extent to which a measurement method or research conclusion appropriately represents what it is intended to measure or establish.

Types of validity

1. Content validity

The extent to which an instrument adequately covers the relevant aspects of a concept.

Example: A nursing examination covering all specified learning objectives.

2. Construct validity

The extent to which an instrument measures the theoretical construct it is intended to measure.

Example: A questionnaire intended to measure anxiety actually reflects anxiety rather than unrelated characteristics.

3. Criterion-related validity

The extent to which a measurement is associated with an appropriate reference or criterion.

Example: Comparing a screening instrument with an accepted reference method.

4. Internal validity

The degree to which a study's design and conduct support the credibility of its conclusions about relationships within the study.

5. External validity

The extent to which findings can reasonably be generalized or transferred to other populations, settings, or conditions.

Reliability

Reliability refers to the consistency or dependability of a measurement under specified conditions.

Types of reliability

  1. Test-retest reliability: Consistency when the same instrument is administered on separate occasions under similar conditions.

  2. Inter-rater reliability: Agreement between different observers or assessors.

  3. Internal consistency: Degree to which items intended to measure the same construct are consistent, commonly assessed using measures such as Cronbach's alpha.

Difference between validity and reliability

Validity

Reliability

Concerns whether a measurement or conclusion is appropriate and accurate for its intended purpose.

Concerns consistency and dependability.

A valid measure should meaningfully measure the intended construct.

A reliable measure produces consistent results under comparable conditions.

Reliability alone does not establish validity.

A measurement can be reliable without measuring the intended construct.

Example: A weighing scale that consistently reads 3 kg above the actual weight may be reliable but not accurate.

Validity and reliability: further points

Face validity

Face validity is the extent to which an instrument appears, on inspection, to measure what it is supposed to measure. It is based on subjective judgement by experts or users, not on theory. For example, if lecturers agree that a test looks like a good test of drug calculation skills, it has face validity. Unlike face validity, content validity is judged against established theory and covers all aspects of the concept.

Validity of an instrument versus validity of a study

  • Instrument validity (face, content, criterion, construct) is about whether the tool measures what it should.

  • Study validity (internal and external) is about whether the conclusions of the whole study are correct and can be generalised.

Factors that affect internal validity

  • Uncontrolled extraneous and confounding variables.

  • Differences between participants (subject variability) and loss of participants during the study.

  • Too little time for data collection or for the intervention to work.

  • A poor, insensitive or inconsistent instrument.

Factors that affect external validity

  • A highly controlled or artificial research environment.

  • Biased selection of participants, for example only males or only one facility.

  • Researcher or experimenter effects on participants.

  • The data collection method.

  • The time when the study was done.

  • A small sample size.

How validity and reliability are improved

  • Review of the instrument by the supervisor and experts.

  • Comparison with other tools whose validity is known.

  • Pre-testing (piloting) the tool on a group similar to the study population, then correcting ambiguous questions.

  • Training research assistants and using standard procedures.

Accuracy and precision

  • Accuracy: how close a measurement is to the true value. It is assessed by comparing with a reference (gold) standard.

  • Precision: how reproducible a measurement is, giving nearly the same value each time it is repeated. It is assessed by repeated measurements.

  • A scale that always reads 3 kg too high is precise but not accurate.

Bias

Bias is a systematic error that can distort research findings.

Types of bias

1. Selection bias

Occurs when the process of selecting participants leads to systematic differences that affect the study's findings.

2. Recall bias

Occurs when participants remember past events or exposures inaccurately or differently across groups.

3. Interviewer bias

Occurs when the researcher's behavior, wording, or expectations influence participants' responses or the collection of data.

4. Measurement bias

Occurs when a measurement method systematically produces inaccurate or distorted results.

5. Publication bias

Occurs when the likelihood of research being published is related to the nature or direction of findings.

Methods of reducing bias

  • Use appropriate sampling methods.

  • Apply consistent eligibility criteria.

  • Use validated instruments.

  • Train data collectors.

  • Standardize procedures.

  • Use blinding where feasible and appropriate.

  • Report limitations transparently.

  • Use appropriate analytical methods.

Data and information

Data

Data are individual observations, measurements, responses, records, or other collected materials used for research.

Examples:

  • Patient age: 24 years.

  • Temperature: 38.2°C.

  • Interview response.

  • Laboratory result.

Information

Information is data that have been processed, organized, or interpreted to convey meaning.

Example:

Data: Individual patient temperatures.

Information: The proportion of patients with fever in a ward during a specified period.

Primary and secondary data

Primary data

Information collected directly by the researcher for the current study.

Examples:

  • Interviews conducted by the researcher.

  • Newly administered questionnaires.

  • Direct observations.

  • Measurements obtained during the study.

Secondary data

Information that was previously collected for another purpose or existing data source and is reused for research.

Examples:

  • Hospital admission registers.

  • Existing patient records.

  • Published datasets.

  • National health reports.

Comparison

Primary data

Secondary data

Collected for the current research purpose.

Previously collected or published.

May be tailored to study objectives.

May have limited relevance or missing variables.

Often requires more time and resources.

May be quicker and less expensive to obtain.

Researcher can plan the collection process.

Data quality depends on the original collection process and available documentation.

More about data

  • Raw (source) data: data as collected, before it has been cleaned, coded or processed.

  • Good data should be complete, consistent, accurate (free of errors) and safely stored.

  • Quantitative data are numbers (for example age or blood pressure). Qualitative data are words, descriptions or categories (for example a mother's experience of labour).

Advantages of primary data

  • Collected specifically for the study's objectives, so it is tailored and up to date.

  • More detailed and complete.

  • The researcher knows exactly how the sample was selected and data collected, so confidence in it can be judged.

Advantages of secondary data

  • Saves time and money because there is no field work.

  • Readily available, for example census reports, the HMIS, patient files, registers and published reports.

  • Not affected by field problems such as bad weather or insecurity.

  • Allows examination of trends over past years, which makes it useful in retrospective studies.

Disadvantages of secondary data

  • Not always accessible.

  • May be incomplete or collected for another purpose.

  • May raise confidentiality issues when using records.

Statistics: definition and branches

Statistics is the science of collecting, organising, analysing and interpreting numerical data. The word also refers to the numerical facts themselves, or to a value that describes a sample (for example a sample mean).

  • Descriptive statistics summarise and describe the collected data using tables, charts and summary measures, without drawing conclusions beyond the data.

  • Inferential statistics use sample results to draw conclusions or make generalisations about the population from which the sample was taken.

Generalizability and transferability

Generalizability

The extent to which quantitative findings can be applied to other populations or settings based on appropriate evidence and study design.

Transferability

In qualitative research, the extent to which readers or users can judge whether findings may be relevant to another context, based on a sufficiently detailed description of the study and setting.

The researcher should not assume that findings from one hospital automatically apply to all hospitals.