Research Methodology

Dr. Samuel Blay Nguah

Kwame Nkrumah University of Science and Technology

2024-05-28

Outline

Methods

  • Qualitative
  • Quantitative
  • Mixed-method

For Each

  • Study designs
  • Sampling
  • Sample size
  • Data collection
  • Data entry tools
  • Analysis methods
  • Presentation of results

What is Research Methodology?

Definition

A structured and scientific approach used to collect, analyze, and interpret quantitative or qualitative data to answer research questions or test hypotheses.\(^1\)

Details

Entails all the important aspects of research:

  • Research design,
  • Data collection methods,
  • Data analysis methods
  • Overall framework within which the research is conducted.\(^1\)

\(^1\) https://paperpal.com/blog/academic-writing-guides/what-is-research-methodology

Questions?

  • Which study methodology should I use?
  • What are the advantages and disadvantages?
  • What are their strengths and weaknesses?

Qualitative methods

What is qualitative research

  1. Involves collecting and analyzing non-numerical data. Example:
    • Interview transcripts
    • Documents
    • Survey responses
    • Images
    • Visuals
    • Videos
    • Audios
  1. Understand concepts, opinions experiences and contents
  2. Used to
    • Gather in-depth insights into a problem
    • Generate new ideas for research \(^1\)
  3. This is more subjective

\(1\) https://www.scribbr.com/methodology/qualitative-research/

Study designs

Two main types:


Population-based:

  • Survey and polls
  • In-depthh Interviews (KII)
  • Focus Group Discussion
  • Observations

Document based

  • Case study
  • Content research
  • Systematic Literature Review

Systematic Literature Review

Definition

This is a comprehensive and unbiased type of literature review that systematically searches, identifies, selects, appraises, and synthesizes research evidence relevant to the research question


  • Desk based study
  • Information already exists
  • There is a specific question to be answered
  • Prone to bias

Systematic Literature Review II

Steps

  • Hypothesis
  • Study question
  • Protocol development
  • Study selection
  • Data extraction
  • Quality assessment

Protocol development

  • Outline problem or knowledge gap
  • Formulate research problem
  • Define search strategy
    • Databases to search
    • Search words
  • Inclusion & exclusion criteria
  • Data extraction techniques
  • Strategy for quality assessment

Systematic Literature Review III

Search strategy

  • Which databases would be used?
  • Which level of evidence will be included?
    • RCTs, etc
  • How will the search be done? “Pubmed with MESH”?
  • What search words to include
  • Quality control: Two independent searches?

Appraisal & synthesis

  • Select studies
  • Extract data
  • Synthesize data


Reporting findings

  • Writing the reviews
  • Publishing

Content research I

Definition

A research tool used to determine the presence of certain words, themes, or concepts within some given qualitative data.

Application

  • Quantify and analyze the presence, meanings, and relationships of certain words, themes, or concepts.
  • E.g: Is there a bias in articles written about Pres. Nana Akufo-Addo?
    • We look specifically for trends, special words that may connote bias or partiality.

Content research II

Uses

  • Identify the intentions, focus, or communication trends
  • Describe attitudinal and behavioral responses to communications
  • Determine the psychological or emotional state of persons or groups
  • Reveal differences in communication content
  • Reveal patterns in communication content
  • Pre-test and improve an intervention or survey before launch
  • Analyze focus group interviews and open-ended questions to complement quantitative data (Mixed method)

Key informant interviews I

Definition

Key Informant Interviews (KIIs) are specialized qualitative interviews conducted with individuals (key informants) recognized for their insider knowledge or unique perspectives on a specific topic.

Key informant interviews III

How?

  • Conducted one-on-one:
    • Face-to-Face
    • Telephone or Online
  • Less formal structured

Why?

  • Provides direct information from stake holders
  • Gives understanding of decisions, etc
  • As a basis for a quantitative study
  • generate recommendations

Key Informant Interviews IV

Procedure

  • Select people who fit criteria
  • Schedule a one-on-one interview session
  • Develop a plan for the interview
  • Be ready to modify on the fly
  • Record responses

Focus Group Discussion

Definition

A qualitative research method that involves facilitating a small group discussion with participants who share common characteristics or experiences that are relevant to the research topic.

Focus Group Discussion II

Planning the session

  1. Plan your session setup
    • Venue
    • How many participants
    • How will they be recruited?
  2. Participants
    • Must resemble target population
    • 8-12 is usual
      • Too few narrows the perspectives
      • Too many means limited perspective from some members
  1. Who will facilitate or lead?
    • Researcher
    • Another person
    • Embedded participant
  2. Who and how will sessions be recorded?
    • Researcher
    • Assistant
    • Video or Audio recording

Focus Group Discussion III

Planning the session

  1. Come up with questions to ask
    • Open-ended questions
    • Should encourage exploration
    • Should generate discussion. Example:
      • What do you think about our services?
      • What do you think about our competitors?
      • What do you wish for in terms oourut services?

Running the session

  • Consent
  • Record with video or sound
  • Make participants comfortable. Everyone introduces themselves.
  • Ask your questions as well as followup questions
  • Engage everyone in the group

Focus Group Discussion IV

Analyzing the data

  • Rewatch the interview
  • Check for themes
  • Understand the patterns emerging
  • Go for further data if required

Validating the conclusions

  • Get someone else to review the data
  • Remember findings are from a small sample size
  • Observer bias - participants know they are being observed so may alter
    • Behaviour
    • Responses

Sampling & Sample size

Sampling population

  • Usually purposive
  • Study type dependent
  • Apply strict inclusion & Exclusion criteria
  • be mindful of saturation
  • 30 participants usually enough

Sampling document

  • Study type specific
  • Literature review: 20-30
  • Others: Subject specific

Data collection & management

Data Collection

  • Ensure you have a good background knowledge
  • Choose the right respondent
  • A good introduction about yourself
  • Consent & confidentiality
  • Minimize disruption
  • Be familiar with your research guide
  • Data collected may include Videos, Audio recording, Write-up, Questionnaires, and Transcripts

Data Management

  • Starts as soon as data collection
  • Revisit data during collection
  • Makes notes during or immediately after interview
  • Organise and store de-identified data
  • Transcribe data - Manual or software
  • Data coding & interpretation:
    • Manual or Software (advisable) e.g. NVIVO

Qualitative Analysis methods - I

Six main methods

  1. Content analysis
  2. Narrative analysis
  3. Discourse analysis
  4. Thematic analysis
  5. Grounded theory’
  6. IPA
  1. Content analysis
    • Examines patterns in words, images, etc
    • Identify frequency of words, phrases, etc
    • Group into frequency, etc
    • Need to approach with a plan e.g. “How many times does ‘happy’ appear in my interview?”
    • Can be time consuming
    • Can miss some important info in the data

Qualitative Analysis methods - II

  1. Narrative analysis
    • Listening to people telling stories and analysing
    • Pay attention to what and how story is told.
    • E.g.: Narrative of a customer about a product
      • Listen
      • Analysimpressionson, likes, hate, etc
    • Disadvantages: Small sample size and poor reproducibility
  1. Discourse analysis
    • Analysis of conversation, interactions, speech, etc within its context
    • Culture and circumstances are important here
    • E.g.: Analyse how a CEO speaks to his employees
    • Should have specific research question beforehand
    • Involves sampling to saturation (New data adds no more information)

Qualitative Analysis methods - III

  1. Thematic analysis
    • Takes often large data
    • Looks for patterns (themes) within data
    • Groups them as such
    • Good for finding out views, experiences and opinions
    • Study question and objective can evolve
    • E.g.: Customers’ opinion about health care provision in a hospital
  1. Grounded theory
    • Uses data to develop theory
    • Go into the analysis with an open mind
    • Develop the theory from ground up
    • E.g.: What what treatment patients prefer and why?
    • This theory develops from the data and not preconceived
    • Useful for an area not researched

Presentation of results

  1. With themes
    • Frequencies
    • Patterns
    • Compare to other studies
    • Make conclusions
  2. As text and narrative

Quantitative methods

What is quantitative research

  • Usually deals with numbers and statistics
  • Used to measure differences or test between groups

Study designs

Descriptive

  • Case study & series
  • Cross-sectional study
  • Qualitative study


Exploratory

  • Cohort study
  • Case control study

Experimental

  • True experimental designs
  • Quasi-experimental designs


Others

  • Systematic Review
  • Meta-analysis

Cross-sectional Study

  • Researcher studies a stratified group of subjects at one point in time
  • Draws conclusions by comparing the characteristics of the stratified groups
  • Well-suited to describing variables and their distribution patterns
  • Can be used for examining associations;
  • Determination of which variables are predictors, and which are outcomes depends on the hypothesis:

Example

Does lead paint ingestion cause hyperactivity OR does hyperactivity lead to lead paint ingestion?

Cross-sectional study

Strengths

  • Fast and inexpensive
  • No loss to follow-up (no follow-up)
  • Ideal for studying prevalence
  • Convenient for examining potential networks of causal links

Weaknesses:

  • Cannot establish causal relationship
  • Does not establish sequence of events)
  • Not practical for studying rare phenomena

Cohort Study

What is a cohort?

A group of individuals who do not yet have the outcome of interest are followed together over time to see who develops the condition

  • Participants are interviewed or observed to determine the presence or absence of certain exposures, risks, or characteristics
  • May be simply descriptive
  • May identify risk by comparing the incidence of specific outcomes in exposed and not exposed participants

Cohort Study

Strengths

  • Powerful strategy for defining incidence and investigating potential causes of an outcome before it occurs
  • Time sequence strengthens inference that the factor may cause the outcome

Weaknesses

  • Expensive, many subjects must be studied to observe outcome of interest
  • Potential confounders: eg, cigarette smoking might confound the association between exercise and CHD

Case-Control Study

Generally retrospective

  • Identify groups
    • Cases (outcome present)
    • Controls (Outcome absent)
    • Matched or unmatched
  • Look backward in time to find differences in:
    • Predictor variables

Assumption:

  • Differences in exposure => Different outcomes

Data collection via:

  • Direct interview
  • Mailed questionnaire
  • Chart review
  • etc

Case-Control Study II

Strengths

  • Rare conditions
  • Short duration & relatively inexpensive
  • High yield of information from relatively few participants
  • Useful for generating hypotheses

Weaknesses

  • Increased susceptibility to bias
    • Separate sampling of cases and controls
    • Retrospective measurement of predictor variables
  • No way to estimate the excess risk of exposure
  • Only one outcome can be studied

Experimental studies

  • To compare 2 or more groups by:
    • Randomization
    • Non-randomization

Assumption:

  • The groups differ solely on the intervention applied
  • Changes from pretest to posttest can be reasonably attributed to the intervention
  • Most basic is the pretest-posttest control group design (RCT)

Experimental studies II

Strengths

  • Controls the influence of confounding variables
  • Randomization eliminates bias
  • Blinding the interventions eliminates bias

Weaknesses

  • Costly in time and money
  • Many research questions are not suitable for experimental designs
  • Usually reserved for more mature research questions
  • Experiments tend to restrict the scope and narrow the study question

Quasi-Experimental studies

  • Aims to establish a cause-and-effect relationship
  • Useful when true experiment cannot be done (ethical, practical, etc)
  • Do not use randomized assignments for comparisons (Non-random assignment)
  • Control group not mandatory

Example

To investigate the relationship between Smoking and immediated Blood pressure measurement

Sampling

Population in research

  • Target population: Any specified group (usually large) of persons, things, or measurement values, e.g. the study population, the sampled population, the target population.
  • Study population: This is a subset of a population, whose properties have been, or are to be, generalized to the larger population or set.
  • Sampling: This is a process of picking a sample from the population.
  • Sample: Members of the study population who are selected for the study. Should be representative of the target population

Sampling in research

Non-Probability

  • Convenience sampling
  • Consecutive sampling
  • Snowballing
  • Quota sampling
  • Purposive sampling
  • Responder driven sampling (RDS)

Probability

  • Simple random sampling
  • Systematic sampling
  • Stratified sampling
  • Cluster sampling
  • Multistage sampling

Simple Random Sampling

Properties

  • Each member of a population has an equal chance of being selected.
  • The sample is chosen randomly without any prior defined selection process.
  • Unbiased method of selecting sample

Simple Random Sampling II

Advantages

  • Minimal knowledge of group required
  • Free from error of classification
  • Suitable for data analysis
  • Free from bias
  • Simple to use

Disadvantages

  • Population should not be dispersed
  • Unusable in an heterogeneous
  • Lacks use of available knowledge concerning population.
  • Need for a well laid out sampling frame.

Systematic Random Sampling

Steps

  • Define your population
  • Generate your sampling interval
  • Randomly select first participant
  • select subsequent regular intervals

Systematic Random Sampling II

Advantages

  • Simple to implement
  • Simple to design
  • No need for a ssmpling frame
  • Good coverage of study area

Disadvantages

  • Might introduce bias
    • Trend
    • Periodicity

Stratified Random sampling

Definition

A method of selecting a sample by dividing a population into smaller subgroups (strata) based on shared characteristics.

Stratified Random Sampling II

Advantages

  • All groups included
  • Statistical precision increased

Disadvantages

  • Strata might be difficult to determine
  • Sampling error difficult to measure

Cluster Sampling

Definition

This is a method of obtaining a representative sample from a population by dividing it into separate groups or clusters.

Steps:

  • Divide population into clusters.
  • Random selection of some clusters
  • All members of cluster included in sample.

Multistage sampling

Convieneince sampling

Definition

Collection of data from population that is convieniently avaiable to provide. Also called “Accidental sampling”

Advantages

  • High participation rate
  • Easy to implement

Disadvantages

  • Difficualt to generalise
  • Bias

Consecutive sampling

Definition

This is a non-probability sampling technique where the researcher selects the sample units from a population in the order in which they appear.

  1. Advantages
    • Fast and easy to carry out
    • Cheap and affordable
    • Researcher selection freedonm
  1. Disadvantages
    • Selection bias
    • Can only be used in small sample sizes

Snowballing Sampling

Sample Size

  • Why bother
    • Identify study population
    • Draw a sample
    • Describe sample (e.g mean)
    • Make inferences about the whole population

Sample size - how many?

  • Determinants
    • Study objectives
    • Type of study
    • Study design
    • Variables to be measured
    • Achievability
  • Too little
    • Not representative enough
    • Cannot make reasonable conclusions about population
    • Waste of resources
    • Ethically improper
  • Too much
    • Waste of resources
    • Data redundancy
    • Ethically improper

Sample size - not one formula fits all!!!

Sample size to determine the average systolic blood pressure of KATH workers


\[ n = \frac{Z^2 \sigma^2}{E^2} \]

Sample size required to determine proportion of hypertensives in KATH



\[ n = \frac{Z_\frac{1}{\alpha}^2 p (1-p)}{d^2} \]

Sample size – Comparing 2 means

  • \(\mu_A\) is the mean in group A
  • \(\mu_B\) is the mean in group B
  • \(n_A\) is the sample size in first group
  • \(n_B\) is the sample size in second group
  • \(\kappa = \frac{n_A}{n_B}\) is the matching ratio
  • \(\alpha\) is the type I error rate
  • \(\beta\) is the type II error

\[n_A=\kappa n_B \;\text{ and }\; n_B=\left(1+\frac{1}{\kappa}\right) \left(\sigma\frac{z_{1-\alpha/2}+z_{1-\beta}}{\mu_A-\mu_B}\right)^2\]

Case-control study

Using the formulas below by Campbell et al \[n=\frac{[Z_{1-\alpha/2}\sqrt{2\bar{p}(1-\bar{p})}+Z_{1-\beta}\sqrt{p_A(1-p_A)+p_B(1-p_B)}]^2}{\delta^2}\] and \[N=\frac{r+1}{2r}\times n\]

  • Where: \(p_A\) and \(p_B\) is the prevalence of the complications in hypertensives and non-hypertensives respectively.
  • \(\delta=p_A-p_B\), and \(\bar{p}=\frac{(p_A+p_B)}{2}\), \(Z_{1-\alpha/2}=1.96\) and \(Z_{1-\beta}=0.84\). \(n\)

Database design, validation and verification

  1. Validation
    • Limits
    • Valid ranges
    • Allowable values
    • Some software better than others
  2. Cleaning
    • Regular review of filled questionnaires
    • Weekly checking of entered data for correctness
  1. Verification
    • Single entry
      • 10% verification
      • Whole database verification
    • Double entry
      • Create identical database
      • Double enter data
      • Picks data entry errors
      • Compare the data from both databases
      • Identify discrepancies
      • Correct errors as necessary

Data Warehousing

Always remember to:

  • Backup the data regularly – 3 copies
  • Backup with versions and dates
  • Keep in the appropriate format
    • Microsoft Excel
    • Text files
    • PDF
    • Tiff

Data migration & cleaning

Cleaning

  • Involves picking out
    • Erroneous and
    • Missing data
  • Picks up
    • Data collection & entry errors
  • Strategy depends on
    • Continuous variable
    • Discrete variable
    • Categorical

Migration

  • Moving data around
  • Should be in stable state
  • Not all software requires this

Data Analysis

Variable types

  1. Independent (predictor) variable
    • Potentially influences, affects or predicts another variable
    • E.g: How age influences income make age the independent variable
  2. Dependent (predicted) variable
    • Potentially predicted, influenced and affected by another variable
    • E.g: How age influences income make income the dependent variable
  3. Software
    • R - Analysis only
    • Microsoft Excel - Entry and analysis
    • Stata - Analysis only
    • SPSS - Entry and analysis

Data analysis

Descriptive analysis

  • Describe your data
  • Categorical variables
    • Univariate
      • Frequencies + percentages
    • Bivariate
      • Frequencies + percentage
  • Numerical variables
    • Normally distributed: Mean(SD)
    • Not normally distributed: Median(IQR)

Inferential analysis

  • Uses sampled data to draw conclusions about a larger population.
  • Allows you to make generalizations about a population based on data from samples.
  • Involves testing hypotheses and deriving estimates
    • P-values
    • Confidence intervals

Presentation of results

  • Text, graphs, maps, tables, etc

Mixed methods

What is mixed-methods research

  • Bring quantitative and qualitative researches together
  • Qualitative component: Used to explore and develop hypothesis
  • Quantitative component: Used to test the hypothesis
  • Assumed to be advantageous to either of the two alone
  • Analysis involves both quantitative and qualitative methods

Mixed methodology - types

5 main types

  1. Convergent parallel design
  2. Explanatory sequential design
  3. Exploratory sequential design
  4. Embedded design
  5. Transformative Design
  6. multiphase design

Convergenet parallel design

  • Simulative collect of both qualitative and Quantitative data
  • Merged but data analysed done seperately
  • Results used to complement each other
  • Advantages:
    • Combines the two data types
    • Quantitative => generalisability
    • Qualitative => provides indepth understanding

E.g: Factors affecting practicing of KMC

Explanatory sequential design

  • Also know as two-phased design
  • Sarts with quantitave and then qualitattive study
  • Qualitative gives indepth knowledge behind the results obtianed in quantitative

E.g: Factors affecting practicing of KMC

  • Quantitative: Determine factors
  • Qualitative: Detemine the thinking behind the observed data

Exploratory qualitative design

  • Similar to above
  • Qualitative done before quantitative
  • Qualitative helps to
    • Design design quantitative methodology
    • Design instrument or questionnnaire

E.g: Factors affecting practicing of KMC

Take home messagge

  • Three types of methodology
    • Qualliative
    • Quantitative
    • Mixed
  • They all have their individual strengths and weakness
  • Know which will best fit your objectives

Any questions? 🤔

Thank you! 🙏

https://sbnguah.github.io