research design
DESCRIPTION
research designTRANSCRIPT
MEANING OF RESEARCH MEANING OF RESEARCH DESIGNDESIGN The research design is the master
plan specifying the methods and
procedures for collecting and
analyzing the needed information.
A set of advanced decisions that
make up the master plan
specifying the methods and
procedures for collecting and
analyzing the needed
information.
PURPOSESPURPOSESProvides a specific blueprint for
conducting a study. It provides maximum control over
extraneous variables and other factors that could influence the results and interfere with the validity of the findings.
The design is the end-result of a series of decisions made by the researcher concerning how the study will be implemented
ELEMENTS-RESEARCH ELEMENTS-RESEARCH DESIGNDESIGNIntervention: To test the effect of
phenomena i.e. what to intervene, procedure, for whom, for how long and the content of intervention
Comparisons : various types -Comparison between two groups-Comparison of a single group at two or more points in time-Comparison of a single group under different circumstances
Groups: Each group in a design is given its own line in the design structure. The major types of assignment to a group are
R = Random assignment N = Non equivalent groups C = Assignment by cut off
Tasks involved in research Tasks involved in research designdesign
DEFINE INFORMATION NEEDED
Design the Exploratory, Descriptive, and/or Causal
Phases of the Research
Specify the Measurement and Scaling Procedures
Construct a Questionnaire
Specify the Sampling Process and the Sample
Size
Develop a Plan of Data Analysis
CAUSALITYCAUSALITY
This describes a relationship
between two variables in which the
presence or absence of one variable
the cause determines the presence
or absence, or the value, of another
variable the effect.
CAUSALITYCAUSALITYA strong correlation must exist
between the proposed cause and the effect.
The proposed cause must occur before the effect.
The cause must always be present
when the effect occurs.
MULTICAUSALITYMULTICAUSALITYThis perspective acknowledges
that more than one variable can be the cause of an effect.
Characteristics of studies examining multicausality include:
Inclusion of more independent variables
than a study based on causality
Complex hypotheses with two or more
independent variables.
PROBABILITYPROBABILITY
This perspective also addresses
the cause and effect relationship
but looks at it from a relative
versus an absolute standpoint.
BiasBias Bias is a systematic deviation from
the truth that distorts the results of research.
Selection biasIntervention biasFollow up biasMeasurement and information bias
MANIPULATIONMANIPULATIONTo manipulate means to move
around or to control the movement.
Control: Control means having the power to direct or manipulate factors to achieve a desired outcome.
INTERVENTIONINTERVENTION
To test the effect of phenomena
i.e. what to intervene, procedure,
for whom, for how long and the
content of intervention
Independent Independent variablesvariablesIndependent variables are those
variables which the researcher has control over and wishes to manipulate.
Dependent variables are those variables that we have little or no direct control over, yet we have a strong interest in.
Extraneous Extraneous variablesvariablesExtraneous variables are those
variables that may have some effect on a dependent variable yet are not independent variables.
Extraneous variables must be controlled through proper experimental design.
Control for extraneous Control for extraneous variablesvariables Researcher must carefully
consider which extraneous variables need to be controlled . E.g. Age, birth weight, diet etc.
Pretest and PosttestPretest and PosttestPretest refers to the
measurement of the dependent variable taken prior to changing the independent variable.
Posttest refers to measuring the dependent variable after changing the independent variable.
DESIGN - TYPESDESIGN - TYPESQuantitative is anything
objective measurable so like your surveys and statistics (numbers)
Qualitative is subjective not measurable like opinions and ideas (words)
Multimethods data in both forms
QUANTITATIVE & QUANTITATIVE & QUALITATIVEQUALITATIVEQualitative
InterpretiveArtisticExplores
meanings
Words
Quantitative
ExperimentsSurveysGenerates facts
Numbers
DESCRIPTIVEDESCRIPTIVE
TYPICAL
TIME DIMENSI
ONAL LONGITUDINAL
CROSS SECTION
CASE STUDY
TRENDCOMPARATIVE
DESCRIPTIVE RESEARCHDESCRIPTIVE RESEARCHDescriptive research is
undertaken to describe answers to questions of who, what, where, when, and how.
Descriptive research is desirable when we wish to project a study’s findings to a larger population, if the study’s sample is representative.
1.TYPICAL DESCRIPTIVEEXAMPLE:
A study to identify the prevalence of cervical cancer
2.TIME DIMENSIONAL: Examines sequences and patterns of change, growth, or trends over time
EXAMPLE:Effect of maternal labor analgesia on
feeding behavior of newbornsMonthly Peer Counselling and the
Juvenile Delinquency
4.Longitudinal: Data collected at more than one point in time over an extended period
Example: To study the changes in professionalism among B.Sc nursing students
5. Cross section: Data collected at one point of time
Example: To study the changes in professionalism among B.Sc nursing students
8.CORRELATIVE 8.CORRELATIVE DESIGNSDESIGNS A correlation is an interrelationship or
association between two variables. Example:Taller people weigh more than shorter
people.1.Descriptive correlative: Describes
the relationship among variablesExample: Coping with disfigurement and length
of hospitalization after head and neck cancer surgery
2.Predictive correlation: Explores what factors have an influence on another variable
EX: A co relational study to predict intentions to obtain Pap smear among urban and rural women
3.Model testing: Testing a hypothesized causal model
4.Multifactorial correlation: Multiple factors influence an outcome
IIDoes the Investigator Influence an
Outcome?• ObservationWatch and RECORD Information• InterventionIntroduces Some Agent to Impact an
Outcome• Natural ExperimentsCurriculum Reform
RETROSPECTIVE: Study begins with DV and looks backward for cause or antecedent
CASE CONTROL: Comparisons of cases with controls
Ex: Factors influencing hypercholesterolemia among cardiac patients
PROSPECTIVE: Study begins with IV and looks forward for effect
EXPERIMENTALEXPERIMENTALAn experiment is defined as
manipulating an independent variable to see how it affects a dependent variable, while also controlling the effects of additional extraneous variables.
Symbols of Experimental Symbols of Experimental DesignDesign
O = measurement of a dependent variable
X = manipulation, or change, of an independent
variableR = random assignment of
to experimental and control groups
E = experimental effect
EXPERIMENTAL EXPERIMENTAL DESIGNSDESIGNS
PRETEST POSTTEST with control
POSTEST ONLY FACTORI
AL
CROSS OVER
RANDOMIZED BLOCK
SOLOMON 4
GROUP
PRETEST POSTTEST
without control
RCT
1. A pretest-posttest design with an experimental and a control group would look like this:
Experimental group O1X O2
Control group O O2Ex Effectiveness of fenugreek
seeds upon hyperglycemia among diabetic clients
2. A pretest-posttest design without an experimental and a control group would look like this:
Experimental group O1X O2
Ex: Effectiveness of laughter therapy upon stress among cancer patients
3. A posttest only design; Data collection at one time after the intervention x o1
x o2
Ex Effectiveness of maternal labor analgesia on feeding behavior of newborns
3. Solomon four group: Data collection before and after intervention for one experimental and one control group, but after only for second experimental and control group
Experimental O1 X O2 O3 control O4 X O5 O6
4.Randomized block : Random assignment to groups within different levels of blocking variable that is not under experimental control
Ex Effect of high fat vs. low fat diet and forced exercise vs. sedentary condition on obesity
5. Cross over: Subjects exposed to all treatments but are randomly assigned to different orderings of treatments ; subjects serve as controls
Ex Effect of music therapy vs. laughter therapy upon stress among cancer patients
6. Factorial design: Experimental manipulation of more than one IV permits a test of main effects for each manipulated variable & interaction effects for combinations of manipulated variables
Ex: Effect of fiber-containing vs. fiber free tube and pectin vs. placebo to reduce diarrhea among critically ill tube fed patients
7. Randomized block design: Random assignment to groups within different levels of a blocking variable not under experimental control ( gender)
QUASI EXPERIMENTAL QUASI EXPERIMENTAL DESIGNDESIGN
NON EQUIVALENT
BEFORE AFTER CONTROL GROUP
NON EQUIVALENT
BEFORE AFTER CONTROL GROUP
NON EQUIVALENT CONTROL GROUP
AFTER ONLY
NON EQUIVALENT CONTROL GROUP
AFTER ONLY
TIME SERIESTIME SERIES
QUASI QUASI EXPERIMENTALEXPERIMENTALCHARACTERISTICS
Manipulation of IVIntervention presentLack randomization
OTHER TYPESOTHER TYPES
CASE STUDYNARRATIVE ANALYSIS
QUALITATIVE OUTCOME ANALYSIS
METASYNTHESIS
DESCRIPTIVE QUALITATIVE
CRITICAL THEORY
FEMINIST
PARTICIPATORY
ACTION
DIMENSIONS – RESEARCH DIMENSIONS – RESEARCH DESIGNDESIGNOBJECTIVE OF STUDY/NATURE OF QUESTIONI EXPLORATORYNeed More Information to Sharpen
QuestionCan’t Measure VariablesII DESCRIPTIVESeeks to Characterize Sets of VariablesAbout People or PhenomenaIII ANALYTICSeeks to Determine Relationships(ultimately casual) Among Variables
TIMEFRAME UNDER INVESTIGATIONI RETROSPECTIVEExamines Background of ResidentsWho Select Primary CareII CROSS SECTIONALA “one shot” Survey Measuring a
Variables(s) at one point in timeIII PROSPECTIVEBegin in the Present and Follow
Subjects Forward in Time
Does the Investigator Influence an Outcome?
I OBSERVATIONWatch and RECORD InformationII INTERVENTIONIntroduces Some Agent to Impact
anIII OUTCOMENatural ExperimentsCurriculum Reform
If you have two or more groups ................Differences
If you have one group and measure the DV two or more times ..................... Differences
If you have one group, and measure the DV once
.....................Relationship
If you have one group, no IV, and observe the DV
once .................................. Descriptive
Which is the Best Research Which is the Best Research Design Design & Method ?& Method ?It depends on the
◦problem of interest, ◦level of information needed, ◦resources ◦researcher’s experience, etc.
TRIANGULATIONTRIANGULATION
It is a strategy in which multiple methods
are used in the research design to study
the phenomenon.
There are five types of triangulation
1.Data triangulation: data from several
sources in a study are examined.
2.Investigator triangulation: Several
investigators with differing backgrounds
examine the same variable.
3 Theoretical triangulation :Several framework are used in a study
4 Methodological triangulation: Several methods such as observation, interviews and questionnaire are used to collect data from the same subjects in a study
5 Analysis triangulation : different techniques are used to analyze the same data.
The over all design The over all design should be such that it:should be such that it:1. Provides for the collection of data in
a manner to produce a high degree
of accuracy, reliability, and validity.
2. Involves a minimum amount of bias and subjectivity on the past of both the research and the respondents being contacted ;
3. Provides the maximum financial economy;
4. Is the most practical approach to the study of the problem
5. The components design fit into each other admirably;
6. Aids in fulfilling the objectives of the study.
Features of good designFeatures of good design
Theory-grounded. Good research
design
finds its roots in theories, that is, it
has the power to test the existing
theories as specific theoretical
expectations are incorporated
in the design.
Situational. Good research design reflects the setting of the investigation.
Feasible. Good research design takes reality into account. The design anticipates potential problems in implementation, measurement and if necessary, includes additional groups or measurements.