HST2122 Health Research Methodology Review for EXAM

Flashcards covering Lectures 1-12 of HST2122 Health Research Methodology with Molly Toomey, covering evidence based practice, research methods, ethics, and statistics.

Evidence-based practice involves what for clinicians?

Clinicians use current best evidence in clinical practice.

Evidence-based practice involves what for educators?

Educators use evidence to engage students, link theory to practice, and inspire future researchers.

What is 'Eminence-based' practice?

Clinical practice based on personal medical clinician preference.

How can you be sure what someone says is evidence-based?

Credentials, referencing, and judging for yourself based on science (empirical grounding) rather than opinion.

What is the benefit of advice based on evidence?

Advice based on evidence has a higher likelihood of success.

Name some different kinds of decision-making that exists.

Decision making by anecdote or (expert) opinion, Decision by press-cutting, and Cost-minimisation.

What is the definition of Research according to Bowling (2009)?

The systematic and rigorous process of enquiry which aims to describe phenomena and to develop and test explanatory concepts and theories.

What is the definition of Research according to Leedy (1989)?

A procedure by which we attempt to find systematically, and with the support of demonstrable fact, the answer to a question or the resolution of a problem.

Why do we do research?

To produce evidence to answer a question.

Research should provide not opinions.

Data.

What is NOT research?

Gathering of facts or information, moving facts from one situation to another, and using it as a buzzword to sell a product.

What is missing from assembly of information that makes it not research?

Regurgitation of other people’s work without interpretation.

What is evidence?

Data used to support (or counter) a scientific theory or hypothesis

Does ‘evidence’ mean we are 100% certain?

No, we can only disprove or strongly suggest.

What characteristic must research questions have?

Research questions must be falsifiable (can be shown to be false).

What is key in science to avoid flukes & be more certain of results?

Replication.

What are the steps in the Scientific Method?

Ask a question, do background research, construct a hypothesis, test your hypothesis, analyze your data, communicate your results.

What are some different types of scholarly work writing?

Research Study (Primary source), Literature Review (secondary source), Theoretical Discussion, Opinion Paper, Program Description, Fiction Writing, & Poetry

What is primary research?

Information gathered through self-conducted research.

What is secondary research?

Information gathered from previously conducted studies.

What are some levels of evidence?

Systematic reviews and meta analyses, Peer-reviewed articles, Government reports, Books, Blogs/websites, Wikipedia.

What is the purpose of a literature review?

Summarize results of previous research to form a foundation, collect ideas on how to gather data, investigate methods of data analysis, study instrumentation & procedures, and assess the success of research designs.

What are the components of article structure?

Title, authors, journal, abstract, introduction, methods/methodology, results, discussion/conclusion, references, acknowledgements, funding, & conflicts of interest.

What are characteristics of Good Research Questions?

Have no simple answer, are researchable, and open the door for other areas of inquiry.

What are characteristics of Bad Research Questions?

Have simple answers, cannot be answered, can only be answered with opinion.

Bad Research Questions often begin with…

who, when, which? how much, how many?

Good Research Questions often begin with…

how, why, what?

Why is it important to reference research reports?

Shows respect, demonstrates knowledge, enables others to locate material, you are not qualified to make statements without it, helps you avoid plagiarism, and it is required.

What are the steps of Hypotheses development?

Develop a Theory, Identify Variables, Identify hypothesis, Test hypothesis.

What is the definition of a hypothesis according to Eric Rogers (1966)?

Single tentative guesses, good hunches – assumed for use in devising theory or planning experiments intended to be given a direct experimental test when possible.

What is the definition of a hypothesis according to Sarantakos, (1993: 1991)?

A tentative explanation of the research problem, a possible outcome of the research, or an educated guess about the research outcome.

What form should hypothesis follow?

Always in declarative sentence form, and they relate, either generally or specifically , variables to variables.

What are the characteristics of a good hypothesis?

It must be able to be tested, it is not a moral/ethical question, it is a prediction of cause and effect, and it is valuable if shown to be false.

How is hypothesis developed?

Comes after the research question, and based on the reading of the literature.

What is qualitative research?

Designed to reveal a target audience’s range of behaviour & the perceptions that drive it with reference to specific topics or issues.

How does qualitative research guide the construction of hypotheses?

Uses in-depth studies of small groups of people to guide & support the construction of hypotheses.

Are the results of qualitative research descriptive or predictive?

Descriptive rather than predictive.

What is quantitative research?

Explaining phenomena by collecting numerical data analyzed using mathematically based methods (statistics).

What is the difference between Qualitative and Quantitative research?

Narrative - Why? Many realities; Measurement - How many? Single reality.

What are popular data collection methods in qualitative research?

In-depth, face-to-face interviews, Focus groups, and Participant observation.

What are popular data collection methods in quantitative research?

Questionnaires, Laboratory experiments, Observational checklists, Field trials, and Longitudinal studies.

What are the differences in explaining phenomena for both quantitative & qualitative research?

Qualitative research inductive (explore, generate ideas from data). Quantitative research deductive (begins with idea [hypothesis], data, draw conclusion based on data…).

What are subjective measures prone to?

Social desirability bias.

What are some Qualitative methodologies?

Ethnography, Case Study, Narrative, Grounded theory, and Phenomenology.

What are four common types of qualitative designs?

Narrative, Grounded Theory, Case Study, and Ethnographic.

How would you describe Narrative research design?

Needing to explore individual stories to describe the lives of people.

How would you describe Grounded Theory research design?

Needing to explore the different experiences of individuals to develop a theory.

How would you describe Case Study research design?

Needing to explore what happens to describe and interpret a case.

How would you describe Ethnographic research design?

Needing to explore a group of people to describe their shared culture.

What are the characteristics of In-depth interviews?

Establishing rapport upfront very important, can read non-verbal cues, see participants’ house, get to understand them better.

What are the characteristics of Telephone interviews?

Requires: Rapport more difficult to establish and lack of Non-verbal cues.

What are the characteristics of focus groups?

More efficient than in-depth interviews, captures group dynamic, participants bounce ideas off each other.

What must facilitator must do during focus groups?

Facilitator must keep participants on track & Homogenous groups work best.

What are two types of Direct observation?

Unstructured and structured observation.

What are some key factors that help ensure Rigour in qualitative research?

Trustworthiness & authenticity, Triangulation, Member check, Bracketing, Prolonged engagement, and Interviewer corroboration.

What are the steps to conduct Qualitative data analysis?

Interviews and focus groups usually recorded (collect data in quiet settings), transcribed into text (very time consuming), then text analysis. After finding themes repeated by different participants with the aim of reduction and interpretation of lots of data into a nice, clear story (narrative). End point: make new theory or hypotheses.

What are the Key learning outcomes for Quantitative Research?

Experimental vs. non-experimental research, Measurement in quantitative research = validity & reliability & Cross sectional vs longitudinal research.

What is Experimental and Non-experimental research?

An Experimental study is where we manipulate something to see the cause of that manipulation. Non-experimental is all other forms of research.

What question do you ask to help define Experimental vs. non-experimental research?

Did the researchers manipulate ANYTHNG in this study?

What should you do if the answer is yes? If no?

If answer is yes = experimental, if answer is no = non-experimental.

To what is Experimental equivalent?

Intervention studies

To what is Non-experimental equivalent?

Observation studies

What are types of validity?

Face validity, Content validity, Construct validity, Criterion or predictive validity.

How do you define Validity?

How well the tool measures what we want it to measure.

How do you define Reliability?

How accurately the tool measures what we want it to measure?

What are 2 kinds of reliability?

Internal consistency and Test re-test.

What are types of Observational/non-experimental studies?

Cross-sectional and longitudinal.

What are Longitudinal studies?

Studies that track participants across several time-points.

What is the Value of longitudinal research?

Correlation vs. causation.

What are the Strengths of cross-sectional research?

Quick and no interventions needed.

What are the Weaknesses of cross-sectional research?

correlation ≠ causation and difficult to tell direction of cause and effect (e.g., overweight kids and TV).

What are Longitudinal studies types?

Prospective (prediction)Study design and measures carefully chosen to test specific hypothesis and Retrospective (not planned)Not intention of original study and measures may have to be adapted.

What is 'Sampling'?

Determining how many will participate in the study and identifying who those individuals will be.

What is representative sampling?

When a researcher selects individuals that are typical of the wider population in key aspects, e.g. SES, gender, age.

What are sampling strategies?

Different approaches for how to select representative samples, such as Probability sampling and Non-probability sampling.

How is generalizability determined?

How well we can generalise study findings from our sample to the population is very much determined by the nature of the sample.

What is the aim of sampling?

Recruit a sample that is as representative of the greater study population as possible & Maximize generalisability.

What is Exhaustive sampling?

Include every single individual in the target population.

What is the definition of Response Rate?

number of people who completed survey / number of people who make up total sample.

What are Types of probability sampling strategies?

Random sampling, Systematic random sampling, Stratified random sampling, Cluster sampling, and Multistage sampling.

What are Types of non-probability sampling strategies?

Convenience sampling, Purposive sampling, and Snowball sampling.

How does probability sampling differ from Nonprobability sampling methods?

Probability sampling provides far better for generalisable results whereas nonprobability sampling methods are more difficult to generalize outside sample to greater population.

How do you describe Sample size in Quantitative studies?

Varies depending on study method and population which researcher hopes to generalise. Researchers conduct a ‘power analysis’ prior to research to give an indication of required sample size sample size calculation provides an indicator of number of participants that are likely required to demonstrate statistically significant difference.

How do you describe Sample size in Qualitative studies?

Varies depending on study method; researchers stop when they reach ‘saturation’ of data.

What is the National Health & Medical Research Council (NHMRC)?

Australia’s version of WHO.

What are Levels in NHMRC hierarchy of evidence?

Systematic reviews, Randomised Controlled Trial (RCT), Pseudo-randomised Comparative Studies, Non-randomised cohorts Studies, Historical groupings, and Single Group / Case Studies.

What is attrition?

Dropout of participants during intervention, potential for only the good students to ‘stick it out’. Lose ability to generalise to all paramed students?

What is Hawthorne effect?

Individuals modify or improve an aspect of their behaviour in response to their awareness of being observed. Blind participants to study conditions wherever possible (ethical implications).

What is Allocation bias?

Differences between groups at baseline. Randomisation into groups should account for this.

What is Detection bias?

Differences in how data is measured between groups.

What is Performance bias?

Other confounding factors contribute to performance at end of experiment (other than intervention alone).

What is Attrition bias?

Group characteristics no longer equal at end of study.

What is Measurement bias?

Outcomes are measured poorly or inaccurately (validity/reliability).

What is GRADE?

Grading quality of evidence and strength of recommendation. Common in the formulation of clinical guidelines. Not practical for individual patients/clinicians/practitioners to make unaided judgements for each decision.

How does GRADE works?

Assess Quality of evidence, Balance between harms and benefits, and Is incremental health benefits worth the cost?

What are GRADE’s Quality of evidence rankings?

High quality, Moderate quality, Low quality, and Very low quality.

How do you define GRADE–Strong recommendation?

We can be reasonably sure that a treatment’s benefits will exceed its costs and associated risks.

How do you define GRADE–Weak (conditional or discretionary) recommendation?

There is (at best) a fine balance between costs and benefits of the treatment, or we cannot be certain about treatment effects.

Examples of Categorical variables

Sex, Age-group, and University attended.

What are examples of Continuous variables?

Age, Time taken to complete a task, and Distance between two points.