RESEARCH QUESTION & HYPOTHESIS

Lt Col A S Kushwaha

RESEARCH• Introduction• The Need for research. Why?• Who all can do research? • What are the types of research?• Where do research ideas come from?• Research Process / cycle• What is a research question?• How to formulate a research question?• Criteria of research question• Hypothesis• Checklist?• Questions if any

INTRODUCTION• “Because we are a poor country, we cannot

afford not to do research”. PM J. L. Nehru

• Health research is not a luxury, to be conducted only by countries with the resources to spare.

“ Keep on going and the chances are that you will stumbleon something, perhaps when you are least expecting it. Ihave never heard of anyone stumbling on something sittingdown.”

(Heath, 1985.)Pasteur said: “In the fields of observation, chancefavours only the prepared mind.”

Science is unpredictable. There is no guarantee thatresearch, actively and methodologically pursued, will leadto the discovery of what it set out to discover. It may do;alternatively, something completely different may be found.

e.g. Minoxidil, Sildenafil, Penicillin (Alexander Fleming)

DEFINITIONS• RE·SEARCH: NOUN: 1. a detailed study of a

subject, especially in order to discover (new) information or reach a (new) understanding. -Cambridge Dictionaries Online, © Cambridge University Press 2003.

• Research requires a question for which no ready answer is available.

HYPOTHESIS

A proposition, or set of propositions, set forth as an explanation for the occurrence of some specified groupOf phenomena, either asserted merely as a provisional conjecture to guide investigation (working hypothesis) or accepted as highly probable in the light of established fact.

THE NEED• Curiosity-driven research• Needs-driven research (Public Health)• Profit-driven research – Ageing population• Participation in collaborative international

research• Publish or perish• Database for comparison• Seed for future research

NEED• The ‘10/90 gap’ • Natural history – CCHF, H1N1• New emerging challenges –H1N1, HIV• Fund allocation / Budgeting • Capacity building – Collab centres• Knowledge generation or translation• Knowledge utilization

A research can be undertaken for two different purposes:

1. To solve a currently existing problem (applied research)

2. To contribute to the general body of knowledge in a particular area of interest (basic/fundamental research)

3. To enhance CV- “Publish or Perish”

PURPOSE OF RESEARCH

WHO ALL CAN DO RESEARCH?

• Students• Teachers• Scientists• Paramedics• Lab workers• Policy Planners• Administration

TYPES OF HEALTH RESEARCH

• Biomedical sciences – Biological, med, clinical, drug, vaccine

• Population sciences – Epid, Demography, Social behavioral

• Health policy sciences – policy, system, services

RESEARCH -TYPES• Multidisciplinary research

• Basic versus applied researchscientific experiments for light (i.e. knowledge)Basic (Not yet applied) and experiments for fruit

– Applied• Quantitative versus qualitative research

RESEARCH IDEAS- WHERE?• Read the medical literature• Attend scientific meetings• Teach—questions asked by students can often give ideas for

research• Acquaint herself/himself with the lines of interest of funding

research organizations;• Develop specific areas of scientific interest• Get new ideas out of her/his own previous research;• Good observer;• Imaginative;• Skeptical attitude - science should be questioned.

RESEARCH QUESTION

Clinical issues and Questions in the Practice of Medicine

ISSUE Question

Normality/abnormality Is a person sick or well?What abnormalities are associated with having a disease?

Diagnosis How accurate are diagnostic tests or strategies used to find a disease?

Risk What factors are associated with an increased likelihood of disease?

Clinical issues and Questions in the Practice of Medicine

ISSUE Question

Prognosis What are the consequences of having a disease?

Treatment How does treatment change the future course of a disease?

Clinical issues and Questions in the Practice of Medicine

ISSUE Question

Prevention Does intervention on people without disease keep disease from arising?Does early detection and treatment improve the course of disease?

Cause What conditions result in disease?What are the pathogenic mechanism of disease?

The literature review is the mother of the Research Question!

WAYS TO SELECT RESEARCH TOPICS

• Personal experience.• Whether you want to evaluate the effectiveness of an

intervention or understand how or why it works• Curiosity about something in the media.• State of knowledge in the field • Solving a problem.• Hot topics under discussion• Personal values• Everyday life.• Gaps in the research and theoretical literature.

(FINER) Criteria for Good Research Question

1. Feasible2. Interesting3. Novel4. Ethical 5. Relevant

Characteristic of a good question

• FEASIBLE– Adequate Number of subjects

• Preliminary sample size• Estimate No available• No who would be excluded or refuse• No that would be lost to follow up

– Most estimates are overly optimistic– Conduct a pilot

Characteristic of a good question

• FEASIBLE

– Technical Expertise : Skills / Eqpt /Experience• Subject recruitment• Measurement of variables• Data analysis

– Cost:• Estimate early• Modify design / abandon question

– Scope• Dont attempt to accomplish too much• Avoid multiple research questions• Narrow the goal and focus on important question

Characteristic of a good question

• Interesting (to the researcher)– Motivations:

• Financial support• Logical or next important step in career building• Getting to the truth of the matter

Characteristic of a good question

• Novel– The question need not be an original

• Can previous observation be replicated• Different study population• Improved technology• Confirmatory study if it avoids the drawbacks of the

previous study

Characteristic of a good question

• Ethical– Check with your IRB– Ensure no physical risk– No invasion of privacy

• Relevant– The most important characteristic– Consider all possible outcomes

• ? How would each affect– Current scientific knowledge– Influence clinical management– Influence health policy– Guide further research

Steps in developing a research question

• Step 1 :- Do not let the research question be forced upon you.

• Step 2 :- Find a general area of interest. ( “ AIDS ” , NCDs”, “MCH” etc ) (basic interest, own clinical observations, discussions with colleagues, med. Conf., Questions asked by our own students)

• Step 3 :- Read “around” the topic in width, (broad & extensive ) not in depth ( intensive )

• Step 4 :- Identify a specific area of interest where gaps in knowledge exist, need to be filled up ( effect of antenatal counselling on Postnatal care)

Steps in developing a research question

• Step 5 :- Read “ into ” the topic– In depth ( intensive )

– Discussions with experts in concerned field.

– Medline search, internet

• Step 6 :- Formulate a tentative research question ( syn. Res. Hypothesis ) : a tentative guess / proposition which we like to prove / disprove

Steps in developing a research question

• Step 7 :- Evaluate the tentative research question for its “suitability” (FINER)

• Step 8:- Make the research question specific

• Step 9:- Write down the Res. Question & its significance

31

Narrowing and Clarifying

• Narrowing, clarifying, and even redefining your questions is essential to the research process.

• Forming the right ‘questions’ should be seen as an iterative process that is informed by reading and doing at all stages.

Defining and refining the research question

HRT use in Postmenopausal Women ------- Endometrial CaDoes post-menopausal hormone replacement therapy

predispose women to develop endometrial cancer?

Is it oestrogen alone or oestrogen in combination with a progestagen? Does the duration of therapy need to be defined as, for example,more than one year?

Should the diagnosis of endometrial cancer be specified asHistologically confirmed?

For the purpose of the research design, the question alsoneeds to be refined. The research will only be able todetermine if there is an association or not. The refinedquestion should therefore be:

Is post-menopausal hormone replacement therapy,as defined, associated with a subsequent increased risk ofendometrial cancer?

The association, if found, will need an explanation, butcannot be taken as meaning causation without furtherquestioning.

If we take another example for a research question,

“Is passive smoking harmful to the foetus?”

the question needs to be better defined and also refined.

“Are the children born to women whosehusbands smoke more than 20 cigarettes a day,of lower birth weight than children born towomen whose husbands do not smoke”?

Problems and solutions

• Vague or inappropriate– Write the question early on– Get specific in the study plan

• How the subjects will be sampled• How will the variables be measured

Problems and solutions

• Not feasible– Too broad

• Reduce number of variables• Narrow the question

– Not enough subjects• Expand the inclusion criteria• Reduce the exclusion criteria• Lengthen the time frame of the study• Use more efficient variables or designs• Add subjects from other sources

Problems and solutions

• Not feasible– Methods inadequate or beyond your skills

• Consult experts• Learn skills / undergo training• Collaborate

– Too expensive• Change study design/ measurements• Seek addl funding

• Ethical uncertainty– Modify the question

Summarising : The research question

1. What topic (idea) of study are you interested in? 2. What has already been done in this area (the literature)? 3. What major outcome(s) (dependent variable) are you interested in? 4. What intervention (independent variable) are you interested in? 5. Are you looking for differences or a relationship (association)? 6. To what group (population) do you wish to apply your results? 7. What is your specific research question? 8. What answer to your question do you expect to find (the research

hypothesis)? 9. Why is this question important today (relevance)?

WHAT TO STUDY?

Describe / report a disease Case report / case series

Study prevalence Survey/cross sectional

Study incidence Cohort

Study causation Cohort/case control

Study associations Cross sectional/ ecological

Study usefulness of diagnostic data Diagnostic test evaluation / clinical prediction rule

Study efficacy of treatmentClinical / field trials

Study outcome and factors influencing it Cohort / economic studies

Case control studiesHealth services evaluation Cross-sectional surveys

40

The Hypothesis Dilemma

• Hypotheses are designed to express relationships between variables. If this is the nature of your question, a hypothesis can add to your research

• If your question is more descriptive or explorative, generating a hypothesis may not be appropriate

41

The Hypothesis Dilemma

A hypothesis may not be appropriate if:

• You do not have a hunch or educated guess about a particular situation

• You do not have a set of defined variables. • Your question centres on phenomenological description • Your question centres on an ethnographic study of a cultural

group • Your aim is to engage in, and research, the process of

collaborative change

• 11.Null and research hypothesis• Null hypothesis (Ho)= Statistical hypothesis;

predict that no relationship exists between variables.

• Research hypothesis(H1)= Alternative hypothesis; state the expected relationship between variables.

Hypothesis are used to state the relationship between two variables and may be stated as

• Null hypotheses (no relationship between two variables).

• Non-directional hypotheses (we don’t know or won’t speculate about the direction of the relationship between two variables).

• Directional hypotheses - We state the direction of the relationship between two variables.

RELATIONSHIPS SPECIFY

• How the value of one variable changes in relation to another.

• May be either positive, negative, or the two variables may not have any relationship to one another.

• Are not necessarily correlations. The type of relationship or association among variables is determined by the level of measurement of each of the two variables.

Examples of relationship type Negative Positive None

1 5 1 1 1 3

2 4 2 2 2 1

3 3 3 3 3 5

4 2 4 4 4 4

5 1 5 5 5 2

BP & Age, HR & age, IQ & Colour of eyes

• Using the previous two examples, the research hypotheses could be as follows.

1. Post-menopausal women who received hormone replacement therapy, of a specified type and duration, are more likely to develop endometrial cancer than post-menopausal women who did not receive such therapy.

2. Children born to women whose husbands smoke more than 20 cigarettes a day are of lower birth weight than children born to women whose husbands do not smoke.

HYPOTHESIS CRITERIA

• Is written in a declarative sentences.

• Is written in the present tense.

• Contains the population.

• Contains the variables.

• Is empirically testable

• The level of significance for rejecting the statistical null hypothesis should always be stated before data are collected. The level of significance usually set at (.05). this means that the researcher is willing to risk being wrong 5% . Generally the aim of the researcher is to reject the null hypothesis because this provides support for the research hypothesis.

THOUGHT FOR THE DAY: WHAT WE WANT TO AVOID

“Your manuscript is both good and original; but the part that is good is not original, and the part that is original is not good.”

Samuel Johnson

Develop a research question & Hypothesis

• General concern – Hb of mother and Birth weight of baby.

RQ - • Is Anemia in pregnancy associated with low birth weight in newborn?

Null Hypothesis• There is no difference in the incidence of LBWs in the mothers who are

anemic and those who are not anemic.

Research Hypothesis• The incidence of LBWs in mothers who are anemic is higher than those

who are not anemic

Just get out and do it!

?

• What are FINER criteria of a good research question?• What is a null hypothesis?• What kinds of relationship may exist between two

variables?• Read into, Read around- which comes first?• What is called as “mother of research”• If we examine “ Why so” in a study- is it a

quantitative or qualitative research?• Incidence is studied by – Descriptive/ Case control/

Cohort/ RCT• Drug or Vaccine for its efficacy is best studied by

which type of study? Descriptive/ Case control/ Cohort/ RCT