research methods ii james kwaku agyen. course outline the course is designed to teach students to...

RESEARCH METHODS II

JAMES KWAKU AGYEN

COURSE OUTLINE• The course is designed to teach students to

formulate- testable research hypotheses- utilize appropriate study designs- collect and analyze quantitative data- and interpret research findings- students will work on practical sessions to enhance

the understanding of topics- to build skills in key techniques including literature

review- Discussions and group activities will be organized to

facilitate students’ interaction

COURSE OBJECTIVES• By the end of this course students will be able to:- Demonstrate an understanding of the range of

epidemiological study designs and their inherentthreats to internal and external validity

- Select the right designs for future research projects- Describe the basic issues related to measurement

of variables- Identify problems with measurement reliability and

validity- Select appropriate analytic techniques for each

study design- Develop students’ skills in formal, scientific writing

•

•

•

•

•

•

•

•

COURSE CONTENTProject planningLiterature reviewVariables and relationshipsNon-experimental including qualitative researchdesigns, experimental research designs; samplingdesigns and proceduresAssigning subjects to study groupsMeasurement concepts and methodsStatistical analysis planReporting research findings

REFERENCES

• Recommended texts• Aday LA and Cornelius LJ Designing and Conducting Health Surveys:

A Comprehensive Guide (2006) 3rd Ed Jossey-Bass, San Francisco

• Ader HJ and Mellenbergh GJ (eds.) Research Methodology inthe Social, Behavioral & Life Sciences (1999) Sage Publications,Thousand Oaks, CA

• Aveyard H. Doing literature review in health and social care.A practical guide (2007). Open University Press, McGraw-HillEducation

• Green Judith and Thorogoo N (Eds) Qualitative Methods for HealthResearch (2004) Sage, London

OBJECTIVESAt the end of this session you should be able to:

• 1. Describe data in terms of frequency distributions,percentages, and proportions.

• 2. Use figures to present data.• 3. Explain the difference between mean, median

and mode.

• 4. Calculate the frequencies, percentages,proportions, ratios, rates, means, medians, andmodes for the major variables in your study thatrequire such calculations.

• 5. Identify other independent variables, if any, that

INTRODUCTION• Variable: a variable is a characteristic that takes on

different values in different persons, places, or things.• Some examples of variables include diastolic blood

pressure, heart rate, the heights of adult males, theweights of preschool children, and the ages of patientsseen in a dental clinic.

• In research, variables are chosen with the assumptionsthat they either;

1. would help us to define our problem (dependentvariables) and its different components

2. or that they were contributory factors to the problem(independent variables) identified.

DNA & JKA

RESEARCH METHODS II

PROJECT PLANNING

•••••

OBJECTIVES

The basic format of all researchResearch planning versus research methodologyThe nature and role of data in researchPlanning for data collectionIdentifying appropriate measurement instruments-Validity and Reliability

• Compare quantitative and qualitative approaches• Ethical issues I research

Research Design• Is a general strategy for solving a research

problem• Provides the overall structure for the procedures

the researcher follows, the data the researchercollects, and the data analyses the researcherconducts

• Research Design – Research planning- this involves identifying your resources, your

procedures, and the forms that your data willtake

- these are done with the central goal of solvingyour research problem in mind

THE BASIC FORMAT OF RESEARCHWhat we must note;

• Research process follows a basic format no matterwhich academic discipline gives rise to it

• That is, general procedure is fundamentally the same- Social science research, nursing research, etc• In planning for a research design, one cannot be

restricted by discipline-specific methodologicalrestraints

- e.g. a nurse may be wanting to solve a problemin her area and may encounter problems that arepsychological

- i.e. researchers must be willing to draw any productive

DO NOT DEPARTMENTALISE KNOWLEDGE

• Researchers do better by generatingproblems from broad generic areaswithin whose boundaries all researchfalls

- Research problems may be generatedfrom; people, things, records, thoughtsand ideas, and dynamics and energy

Categories of Research Problems• People: problems related to children, adults,

families, communities, cultural groups, etc.• Things: problems related to animal and vegetable

life, viruses and bacteria, inanimate objects,matter, stars, galaxies, etc.

• Records: problems related to newspapers,journals, letters, minutes, legal documents,census reports, etc.

• Thoughts & Ideas: problems related to concepts,theories, perceptions, opinions, beliefs, etc.

• Dynamics & Energy: problems related to humaninteractions, metabolism, quantum mechanics,

4-13

1. A question is posed- In the mind of the researcher, a question arises that

has no known resolution2. it’s a matter of words- The researcher converts the question to a clearly

stated research problem

3. it’s worth a guess- The researcher poses a temporary hypothesis or series

of hypotheses4. The search is on

- The researcher searches the literature for ideas thatshed light on the problem and for strategies that may

THE BASIC FORMAT OF RESEARCH

THE BASIC FORMAT OF RESEARCH5. Data

- The researcher collects data that potentially relate tothe problem

6. How do the data fit together?- The researcher arranges the data into a logical

organizational structure.7. The data speaks

- The researcher analyzes and interprets the data to determine their meaning

8. Data supports or not- Either the data seemingly resolve the research

problem or they do not. Either they support the hypotheses or they do not

Research Planning verses Research Methodology

• Research Planning: the general approach to planning a research study; is similar across disciplines

- Data vary in nature. One cannot deal the same way with malaria in children as he/she will do with a historic document

Research Planning v. Research Methodology

Therefore, in planning the research design

- consider the kind of data the study will require

- choosing viable research problem, reasonable way of collecting and interpreting the data

• Research Methodology- the techniques one uses to collect and

analyze data; may be specific to a one’s discipline

• Universality: - the research project should be one that

might be carried out by any competent person.

- The researcher is a catalyst who collects, organizes, and report what the collected data seem to indicate.

General Criteria for a Research Project

4-17

• Replication:- the research should be repeatable;- any other competent person should be

able to take the problem and, collecting data under the same circumstances and within the same parameters you have used, achieve results comparable to yours.


• Control: - the researcher must isolate, or

control, factors that are central to the research problem; control is important for replication and consistency within the research design.

• Measurement: the data should be able to be measured in some way


The Nature and Role of Data in Research• Data are transient and ever changing.

• Data are not absolute reality but manifestations ofreality.

• Data is primary or secondary.- Primary data is the layer closest to the truth.- Secondary data are derived, not from the truth,

but from primary data.• Data must meet certain criteria to be admitted

to study; any data not meeting the criteria areexcluded from the study. Defective data mayaffect the validity of the research conclusions

• The restrictions used to admit certain data for

Planning for Data Collection

Answers to the following questions will bring theprocess into focus.

1.

2.

3.

4.

What data are needed?

Where are the data located?

How will the data be obtained?

How will the data be interpreted?

4-19

Identifying Appropriate MeasurementInstruments

• We pin down data by measuring it in some way.• Measurement instruments provide a basis on

which the entire research effort rests.• A research effort employing faulty measurement

tools is of little value in solving the problem underinvestigation.

• In planning the research project, the nature ofthe measurement instruments should be clearlyidentified.

• Instrumentation should be described in explicit,concrete terms.

• Instruments should have 4-20 a reasonable degree of

• Measurement is limiting the data: this meansby a certain standard, we set limit to constrainthe data; what is foot, a mile, a pound. They areeach a unit of measure governed by a numericalconstrain; so 12 inches constrains a foot, 5,280

Measurement as a Tool of Research• Measurement: Is limiting the data of any

phenomenon—substantial or insubstantial—so that those data may be interpretedand, ultimately, compared to a particularqualitative or quantitative standard.

- Phrases in the definition;

• “of any phenomenon-substantial orinsubstantial”: this simply means that there isnothing a researcher cannot measure.

- Substantial measurements those thingsbeing measured have physical substance,e.g. weight of a baby, diameter of a wire

- Insubstantial measurements these exist onlyas concepts, ideas, opinions, feelings, orother intangible entities. e.g. the degree towhich student have learned, the extent to

Phrases in the definition

Four Scales of Measurement:

• Nominal

• Ordinal

• Interval

• RatioLeedy & OrmrodPractical Research: Planning and Design, 10e© 2013 Pearson Education, Inc. All rights reserved.

4-23

Nominal Scale of Measurement

• Measures data by assigning names

• Things can be measured nominally in an infinitenumber of ways

• Simplistic

• Divides data into discrete categories

• Statistical procedures = mode, percentage,chi-square test

Leedy & OrmrodPractical Research: Planning and Design, 10e© 2013 Pearson Education, Inc. All rights reserved.

4-24

Ordinal Scale of Measurement• Think in terms of symbols (>; <) that is we can

compare various pieces of data in terms of onebeing greater than the other

• This scale allows to rank-order data, thus the nameordinal. NB: Difference in ranks is meaningless

- e.g. level of education can be measured on theordinal scale by unschooled, elementary, highschool, college or graduate education

- Lecturers can be measured as assistant lecturer,lecturer, senior lecturer, associate professor, andprofessor

• Statistical procedures used; median, percentile

Interval Scale of MeasurementThis scale of measurement is characterized by twofeatures:1. Has equal units of measurement, as on the

number line but 2x does not mean twice x on thescale

2. Zero point established arbitrarily (zero doesnot mean the total absence of the quality beingmeasured) e.g. temperature measurement scalessuch as the Fahrenheit (F) and Celsius (C)

• Rating scales, such as those used in surveys, areinterval scales.

• Statistical procedures = means, standard

Ratio Scale of Measurement• Characterized by equal measurement units(similar to an interval scale)

• Has an absolute zero point (zero means totalabsence of the quality being measured) and”minus” distance is not possible

• Can express values in terms of multiples andfractional parts, therefore, 2x means twice x. e.g.

a yard is a multiple i.e. 36 times a 1-inch distance; aninch is one twelfth, a fraction part, of a foot

• Ratios are true ratios (Yardstick, feet) e.g. 36:1 and1:12. it allows us to make comparisons that involvemultiplication or division

Summary of the Scales of Measurement• Nominal scale: One object is different from another.

• Ordinal scale: One object is bigger or better or moreof anything than another.

• Interval scale: One object is so many units (degrees,inches) more than another, but zero does not mean

total absence of the quality in measurement

• Ratio scale: One object is so many times as big orbright or tall or heavy as another, zero means totalabsence of the quality in measurement

Validity & Reliability of Measurement

• Validity = the extent to which ameasurement instrument measures whatit is intended to measure.

• Reliability = the consistency with which ameasurement instrument yields a certainresult when the entity being measuredhas not changed.


4-30

Validity of Measurement Instruments• Face Validity: the extent to which an instrument looks like

it’ s measuring a particular characteristic; relies onsubjective judgment.

• Content Validity: the extent to which a measurementinstrument is a representative sample of the content areabeing measured. E.g. If it reflects various parts of thecontentbehaviors

domain in appropriate proportions and requirescentral to that domain. Example assessing peoples

achievement in some area

• Criterion Validity: the extent to which the results of anassessment correlate with another, related measure. E.g.salesperson’s effectiveness

• Construct Validity: the extent to which an instrumentmeasures a characteristic that 4-31 cannot be directly observed

•

•

•

Ways to Demonstrate the Validityof a Measurement InstrumentTable of specifications: the researcher constructs atwo-dimensional grid listing the specific topics andbehaviors that reflect achievement in the domain.

Multitrait-multimethod approach: two or more differentcharacteristics are each measured using two or moredifferent approaches. The two measures of the samecharacteristic should be highly related.

Judgment by a panel of experts: several experts in aparticular area are asked to scrutinize an instrument toto ascertain its validity for measuring the characteristic inquestion.


4-32

Reliability

• Reliability is the consistency with whicha measuring instrument yields a certainresult when the entity being measuredhasn’t changed.

• Instruments designed to measuresocial and psychological characteristics(insubstantial phenomena) tend to beeven less reliable than those designedto measure physical (substantial)phenomena.


4-33

•

•

•

•

Determining the Reliability of aMeasurement Instrument

Interrater reliability: the extent to which two or moreindividuals evaluating the same product or performancegive identical judgments.Internal consistency reliability: the extent to which all ofthe items within a single instrument yield similar results.

Equivalent forms reliability: the extent to which twodifferent versions of the same instrument yield similarresults.

Test-retest reliability: the extent to which a singleinstrument yields the same results for the same peopleon two different occasions.


4-34

Quantitative vs. Qualitative Approaches

• Quantitative Research: involves looking at amounts, orquantities, of one or more variables of interest.Researchers attempt to measure variables in some way.

• Qualitative Research: involves looking at characteristics, orqualities, that cannot easily be reduced to numericalvalues. Researchers attempt to examine nuances andcomplexities of a particular phenomenon.

• Quantitative and Qualitative Processes▪ formation of one or more hypotheses▪ review of the related literature▪ collection and analysis of dataLeedy & OrmrodPractical Research: Planning and Design, 10e© 2013 Pearson Education, Inc. All rights reserved.

4-35

Quantitative vs Qualitative ResearchA Comparison

• Purpose

• quantitative -- seeks explanations and predictions that aregeneralizable. To establish, confirm, or validate relationshipsand to develop generalizations that contribute to existingtheories

• qualitative -- seeks better understanding of complexsituations.

Process

• quantitative -- methods defined before study begins; allowsobjective measurement; researchers remain detached

• qualitative -- holistic/emergent; researchers remain openand immerse themselves in the complexity of the situationand interact with participants; variables emerge from data.

Quantitative vs. Qualitative ResearchA Comparison

• Data Collection

• quantitative -- data collected in form that is easilyconverted to numbers

• qualitative -- researcher is the research instrument;verbal and nonverbal data collected. They operateunder the assumption that reality is not easily divideddiscrete measurable variables

Reporting Findings

• quantitative -- data reduced to averages; scientificstyles of reporting

• qualitative -- construct interpretive narratives from data

Combining Quantitative and QualitativeDesigns

• Mixed-method design provides amore complete picture of a particularphenomenon than either approach coulddo alone

Qualitative or Quantitative?Guidelines

Consider your own comfort with the assumptions of thequalitative tradition.

Consider the audience for your study.Consider the nature of the research question.Consider the extensiveness of the related literature.Consider the depth of what you wish to discover.Consider the amount of time you have available for conductingthe study.Consider the extent to which you are willing to interact withthe people in your study.Consider the extent to which you feel comfortable workingwithout much structure.Consider your ability to organize and draw inferences from alarge body of information.

Considering the Validityof Your Method

The validity of the research project as a whole meansits accuracy, meaningfulness, and credibility.

Researchers should consider two questions:1. Does the study have sufficient controls to ensure

that the conclusions drawn are truly warranted bythe data?

2. Can the results obtained reasonably be used tomake generalizations about the world beyond thatspecific research context?

The answers to these two questions address the issuesof internal validity and external validity, respectively.

Internal vs External Validity

• Internal Validity: the extent to which the design anddata of a research study allow the researcher to drawaccurate conclusions about cause-and-effect and otherrelationships within the data.

• External Validity: the extent to which the results of aresearch study apply to situations beyond the studyitself; the extent to which conclusions can be

generalized.

Strategies to Increase Internal Validity

• Conduct a controlled laboratory study

• Conduct a double-blind experiment

• Use unobtrusive measures

• Use triangulation

Strategies to Increase External Validity

• Conduct the study in a real-life setting.

• Use a representative sample.

• Replicate the study in a different context.

Validity in Qualitative Research

• Spend extensive time in the field. Spend extensivetime in the field

• Conduct a negative case analysis. Contradict existinghypotheses, then continually revise them until allcases are accounted for

• Use thick description. Readers able to draw their ownconclusion

• Seek feedback from others.

Ethical Issues in Research Assurances

• Honesty with professional colleagues

• Internal Review Board (IRB)

• Protection from harm

• Informed consent

• Right to privacy

• Professional code of ethics

The Value of a Pilot Study

Pilot study: a brief exploratory investigationbefore the main study

– to try out particular procedures, measurementinstruments, or methods of analysis.

– to determine the feasibility of the study.– to identify what approaches will and will not

be effective in solving the overall researchproblem.

THANK YOU

research methods ii james kwaku agyen. course outline the course is designed to teach students to...

Documents