Please enable JavaScript.
Coggle requires JavaScript to display documents.
PSY 390 STUDY UNIT 3: RESEARCH DESIGNS AND STRATEGIES (Chapter 8: Non-…
PSY 390 STUDY UNIT 3: RESEARCH DESIGNS AND STRATEGIES
Chapter 7: Experimental & Quasi- Experimental Research Designs
Research Strategies, Research Designs and Research Procedures
According to Gravetter & Forzano (2016), a research study undergoes three stages namely
determining a research strategy
determining a research design
determining research procedures.
2. determining a research design
A research design addresses
how to implement the strategy.
It refers to the plan, general framework, or procedural arrangement to be used for seeking an answer to a research question.
It is
concerned with the selection of units (types and levels of variable) and the arrangement of conditions of a study
so that the relationships among variables can be examined
3. research procedures.
involves filling in the details that precisely
define how the study is to be done.
1. determining a research strategy
A research strategy refers to the
general approach and goals of a research
study, and is usually determined by the kind of question the researcher hopes to address.
Experimental Research Strategy
The
experimental
research strategy is used to
answer cause and effect questions
about the relationship between two variables.
An experiment attempts to establish a cause-and-effect relationship
by demonstrating that changes in one variable are directly responsible for changes in another variable.
Basic concept
Experimental Group vs. Control Group
The
experimental group
is
exposed to
the
independent variable
-
The
control group
is not subjected to the independent variable, and is u
sed for comparison
with the treatment group.
Treatment group/ Experimental Group
Treatment is another term for the independent
variable that is manipulated by the researcher. The experimental group is therefore sometimes called the treatment group
Placebo
Placebo
is a
substance
(known to have no pharmacological effect) that is a
dministered to the subjects who believe that it has therapeutic effects.
Findings indicate that many patients suffering from various illnesses show improvement after receiving a placebo (in the forms of pills, solutions, or injection). This phenomenon is known as the placebo effect.
Extraneous Variables
They are variables who have
effects on the dependent variables but are not themselves being measured
or examine like the independent variable.
Their effects on the dependent variables are undesired sources of variation for the investigation of the relation between the independent and dependent variables.
They are variables
to be controlled for in an experiment
.
In the investigation of whether self-esteem affects academic achievement, for instance, extraneous variables include parental expectations, teaching methods, socio-economic background and all other variables that may affect the dependent variable (academic achievement).
Independent vs. Dependent variables.
Variables are entities
(either qualitative or quantitative)
which interact
to produce the research phenomenon that is being measured (i.e. may take two or more values).
Dependent
variables in the research context are variables which are affected by other factors. They are usually the
variables that researchers want to measure
, investigate and understand.
.
Independent
variables are
variables that impact upon the dependent variables
. They are considered the cause of the dependent variables, and their relationships with the dependent variables are under the researcher’s investigation.
For example, a researcher hypothesizes that exposure to frightening TV programmes causes children to have nightmares. The independent variable is the amount of exposure to frightening TV programmes. The dependent variable is the frequency of reported nightmares within a particular period after watching the film.
Hawthorne Effect
Hawthorne effect is a
reaction of the participants to the researcher’s observation
. (i.e. participants behave in a different way under observation)
In the Hawthorne plant studies, researchers examined the effects of various factors, including the amount of lighting, rest period and quitting time, on work performance. In the study on the amount of lighting, one group of workers worked under increasing levels of illumination and another group under constant illumination. Contrary to expectation, both groups of workers had an increased productivity of similar magnitude. What puzzled the researchers was why different conditions resulted in similar results. Other strange findings included an increase in productivity when rest periods were eliminated and later quitting times were instituted. A plausible explanation for the counterintuitive results is that workers reacted not only to the manipulation of the experimental treatment, but also to being observed. The special attention the researchers gave them motivated them to work better than before, irrespective of the treatment they received.
Pre-tests vs. Post-tests
Pre-tests
is the
measurement of dependent variables before the introduction of the treatment
(or exposure to the effect of independent variables). They are supposed to present the original conditions of an event
Post-tests
is the
measurement of the dependent variables after the treatment
. They are supposed to record the effect of the independent variables on the dependent variables.
Establishing Causation
Causal inference
is one of the most important aims for scientific investigation, because finding out the causes of events or phenomena
enhances our understanding of things that happen
around us.
Strategies for Establishing Causality
Comparison.
‘To compare’ is to identify similarities and differences among different events. Such information is important for logical reasoning about the sequences of events, and estimation of effects.
(i.e. increases inferential strgenth)
Control.
The major function of control is to
isolate the supposedly causally related variables
from the influence of other irrelevant variables (i.e. extraneous variables), so that their relations can be accurately identified.
Manipulation.
The researcher
changes
the value of one
variable
to create a set of two or more treatment conditions.
Measurement.
A second variable is measured for a group of participants to obtain a set of scores in each treatment condition.
INTERNAL AND EXTERNAL VALIDITY
Internal validity
is
concerned with
the extent to which a
causal claim(s)
between the independent and dependent variables is appropriate or valid.
An
internally valid study provides strong evidence of cause and effect
.
.
External validity
concerns the extent to which results obtained in a research study hold true outside the constraints of the study.
(I.e. genralisability)
Any
factor that limits the ability to generalize
the results from a research study is considered a
threat to external validity.
Internal and external validity. Relationship
the strategies for maximizing internal validity limit the external validity by excluding different settings, times, and people. In other words, the more control is used, the more difficult it might be to generalize the results to other settings, times, and people.
inversely related
TYPES OF EXPERIMENTAL DESIGNS
Laboratory vs. Field Experiments
Laboratory experiments
These are conducted in a laboratory, and external factors are controlled.
(High internal validity)
Lab experiments used in social science resemble those used in natural science.
However, some social scientists argue that people’s behaviour in a confined laboratory may not be the same as those in their natural environment, and therefore some researchers prefer to conduct the experiment in the ‘field’.
(Low external validity)
Field experiments
These are conducted in natural situations (e.g. schools, housing estates, companies). In the field, researchers may find it more difficult to control some situational factors than in the laboratory.
(Low internal validity)
However, field experiments enjoy
higher external validity
than laboratory experiments as phenomenon are naturally occuring and are not subected to artifical intervention or control
True vs. Quasi-Experiments
True experiments
true experiments are
characterized by a random assignment
of subjects, and they use data from multiple groups to
compare independent variables.
As true experiments are research designs that
lead to unambiguous causal inference
, they need to apply the major principles that safeguard high internal validity.
Characteristics of True Experimental Designs
The following are three major characteristics of true experiments
Control of Variables
The experimenter has a
high degree of control
over the assignments of subjects, the arrangement of experimental conditions,
systematic manipulation of independent variables
, and choice of dependent variables.
Control serves to
exclude threats to internal validity
. Some forms of control include:
Varying variables systematically
changes variables in a particular way so that their effects can be recorded with respect to different changes.
This is also known as experimental manipulation. Sometimes, variables are allowed to change but their effects are balanced out among groups by randomization.
Theoretically, using this procedure should
generate only random errors, not systematic bias
, and thus no directional changes should be introduced to the dependent variables.
Assigning subjects to different conditions
is when the experimenter assigns subjects to conditions rather than observing them in naturally occurring conditions.
Assignment of subjects is a unique and most critical defining characteristic
that differentiates experiments from non-experiments. No subject assignment is carried out in non-experimental designs.
Holding variables constant
keeps variables unchanged throughout the experiment.
By holding some extraneous variables (variables not of direct research interest) constant, the researchers can
prevent them from contaminating the research results.
Removing the effects of irrelevant variable
s by statistical means is achieved by specific statistical methods such as regression. (i.e. removing outliers)
Intervention or Treatment of independent variables
Intervention or Treatment of independent variables Some type of intervention or treatment of independent variables is implemented in true experiments.
The
independent variables
are
systematically manipulated
so that their effects can be observed. In a way, the threat of reverse causation can be eliminated.
Meaningful Comparison
The
purpose of control
over various arrangements is basically to
establish a proper comparison to evaluate the effectiveness
of the treatment.
The comparison should be made between two or more groups that are treated differently as regards to independent variable levels.
Types of True Experimental Designs
Factorial Designs:
There is
more than one independent variable
in the designs and the design is so structured that every level of an independent variable is associated with every level of another independent variable.
For instance, if the
two independent variables
are sound level (high vs. low) and task difficulty (difficult vs. easy), then the design
should consist of four treatment groups
with different combinations of the independent variable levels: "high sound level and difficult task", "high sound level and easy task", "low sound level and difficult task", and "low sound level and easy task".
A major
strength
of factorial designs lies in their
ability to examine more than one independent variable at a time
. Interaction between the examined independent variables can also be investigated.
Weakness
Due to the investigations of
more independent variables, more subjects
are required compared to other simpler designs.
Within-Subject Design:
uses
a single group of participants
, and tests or observes each individual in all of the different treatments being compared.
This means that the same group of individuals participates in every level of the independent variable so that each participant experiences all of the different levels of the independent variable.
Between-Subject Design:
compares different groups
of individuals, where the researcher manipulates the independent variable to create different treatment conditions and a separate group of
participants is assigned to different treatment conditions.
The dependent variable is then measured for each individual and the researcher
examines
the data, looking for
differences between the groups.
Quasi-experiments
Quasi-experiments aim to examine causal relationships and follow the three basic principles of experimentation:
Control,
Assignment of subjects and
Comparison.
However, quasi-experiments are
not true experiments because there is no random assignment
of subjects and/or no multiple groups.
Types of Quasi-Experimental Designs
Two major types of quasi-experimental design are:
non-equivalent control
group designs
has a control group but no randomization of subjects. It suffers mainly from group threats, including selection threat, regression to the mean, and the selection-by-time interaction
interrupted time series
designs
consists of a single group with multiple measures. It is susceptible to time threats such as history, maturation, test reactivity, and measurement decay
Establishing Internal Validity in Quasi-Experiments
Consulting additional information or examining
additional data
in relation to threat.
For instance, the literature or additional data gathered might inform the researchers that gender is unrelated to the studied phenomenon, which removes one selection threat (gender difference) as a possible rival explanation.
Using reasoning that is based on theory
or common sense to exclude a particular threat as an alternative explanation.
For instance, it is reasonable to rule out maturation threat to a quasi-experiment that takes two measures on adults’ metabolic rates within one week, because the metabolic rate is generally quite stable for adults within a short period.
Special design features
that may check or exclude various threats (e.g. reverse causation, time or group threats) to internal validity.
For instance, the use of pre-tests
in quasi-experimental designs can strengthen the causal inferences because any pre-treatment differences between non-equivalent groups can be checked. As discussed above,
researchers may
also use multiple measures
to identify any trends or cycles that exists before the treatment, and subsequently excludes them when estimating the effects of the independent variables.
Chapter 8: Non- Experimental Research Designs
Introduction
Non-experimental designs, in principle, cannot answer causal questions.
Instead, they are
used to answer descriptive, associative and comparative questions.
Generally, they are
simpler than experimental designs
in the number of groups used, number of measurements, and types of statistical analysis. In fact, all nonexperimental research designs are
characterized by their lack of artificially introduced intervention, control groups and randomization of subject assignments.
Types of Non- Experimental Research Strategy
Surveys and Cross-Sectional Studies
Answers Comparative questions: Differences between groups (also descriptive and associative)
Characteristics
A survey or cross-section study involves the
comparison of the same variable(s) of two or more groups (or cases)
.
In cross-sectional comparison designs, the
measurements
of more than one case
take place at the same time
so as to eliminate errors due to different time sampling (i.e. observations made at different time spots).
Administration of Surveys
To conduct a survey, the survey questionnaire must be designed to meet the
objectives of the survey.
The survey questionnaire can be administered to potential respondents through
face-to-face contacts
through telephone calls
or through using the computer (such as email surveys or online surveys.
careful consideration in designing the questionnaire.
Benefits and Limitations of Surveys and Cross-Sectional Studies
Benefits
The survey design is
simple
and easy to administer.
Cross-sectional designs are frequently used in field studies where researchers have little control over subjects. This approach is
useful in determining whether two or more variables are similar or different
.
Limitation
A limitation is that it is
difficult to identify the causes
for the group differences. Differences among cases may be due to the effect of some external variables. But it may also be due to the inherent group differences before exposure to the variables. Therefore,
no causal conclusion can be drawn
from cross-sectional comparison design.
Trends and Longitudinal Studies
Answers Comparative questions : Changes with time (also descriptive and associative)
Aims of Longitudinal Studies
Assess the need for changes of status.
To
assess changes over time in order to modify policy
, treatment, methods, etc. so that resources can be effectively allocated at various stages to meet the need for the changes.
For example, different doses of medicine are given to chronically ill patients with reference to the assessment of their physical conditions over a period.
Discern Trends and patterns
To discern trends in order to predict what will happen in the future.
Capture Developmental changes.
An example is the developmental changes of children over time, such as their growth in weight, height, perceptual and cognitive ability, social ability, personality, etc. Psychologists and educational experts want to
identify
the trend of
development in order to develop
educational programmes, invention of suitable games and toys for children of different age groups, or prescription of
appropriate
diet for their physical and growth needs.
Benefits and Limitations of Longitudinal Studies
Benefits
Provides a wealth of in depth data
Limitation
Longitudinal studies are usually more
costly
(in time and resources) to conduct than case studies and cross-sectional studies.
The subject
mortality rate is also relatively higher,
because research participants may have moved house, changed their employment status, or just lost interest over a long period.
Data analyses are also more complex
to conduct.
No causal conclusion
Characteristics of Trends and Longitudinal Studies
A longitudinal study involves
studying the same group
two or more times
over a prolonged period
.
the same group is measured at different times, and comparison is made between the measurements.
it attempts to
identify changes of its characteristics over time.
Simple Case Studies
Answers Descriptive, associative question
Aims of Case Studies
Exploration of social phenomenon.
To
identify variables that are relevant to the research topic
and explore the relationships among different variables (specifically, whether the variables are associated with one another, and in what ways are they associated).
Preliminary exploration is helpful in sketching a picture for formulating research hypothesis.
Initial testing of a hypothesis.
To pilot-test research hypothesis
so that more efficient use of resources can be carried out
in subsequent actual research.
However, such pilot tests are limited to tests for associations, not causal relationships.
Description of unique social phenomena
.
The goal is primarily to
obtain in-depth understanding
of a particular group or
unique phenomenon
,
not for generalization
, though insights might also be made about other groups.
Characteristics of Case Studies
Only one case is studied
The most important characteristic of case studies is that the
measurement is carried out only once
. (only a single case is studied)
No attempt
is made
to make comparisons
between groups (between studied groups, or within the group at different times). When two or more variables are studied, their associations can also be explored.
No causal claims
case study
cannot confirm researcher’s hypothetical causal claims
, because case studies are liable to threats to internal validity. There is no control for extraneous variables, so it is difficult to refute other possible explanations.
Benefits and Limitations of Case
Case studies are the basic building blocks of other research designs. All other research designs use one or more case studies as their unit of analysis.
This design is popular among social scientists because of its
simplicity and ease of application
. It enables them to identify relevant variables and associations among variables, which helps plan for future research.
For some researchers, more focused attention to a particular group or phenomenon can be arranged in case studies. However, its use is
limited to studying correlational relationships, not causal relationships.
Longitudinal Comparison Studies
Questions Comparative: Changes with time and differences between groups (also descriptive and associative)
Characteristics of Trends and Longitudinal comparison Studies/Panel studies
It is concerned with the question, ‘Are P and Q different over time?’ or ‘Does the similarity (or difference) change over time?’
involves the
measurement and comparison of two cases/groups
at two time spots, so that the relative changes that occur in the two cases can be estimated. It fulfils the general functions of both cross-sectional and longitudinal studies discussed above.
Evaluation of Longitudinal Comparison Studies
While such studies provide much information
(Strong correlation and high external validity)
for different comparisons in data analysis, their
design is relatively complicated
and is liable to a relatively
higher mortality rate
.
This is because when two groups are involved in such comparative studies, data collection and comparisons will be adversely affected when one group loses interest and refuses to participate in the study anymore, or becomes unavailable due to death or loss in contact, etc.
Since what happens between the two time points is
beyond the researcher's control
, many factors could have happened to affect the results (differences). Therefore,
it cannot confirm causal conclusions.
Observational Studies
Observational studies
involve an observer seeing what is happening about a phenomenon
. Restricted to the study of present and observable events.
Benefits
Its major advantages are to
allow researchers to study people who are not able to fully express themselves
, and
phenomena
that the participants are
not fully aware of
.
Limitation
There are however problems like
difficulties of access to observation
reactivity from the observed, and
the questionable
ethical practices
of some studies.
Observational data can be recorded by several means, including using observational schedules and field notes. Researchers need to
balance the roles of observation and participation to achieve their research goals
.
Observers should be accurate in observation, be able to think and write quickly and have a good memory.
Major errors and bias in observations are caused by the inadequacies of human perception, theoretical and technical problems, and interaction between the observer and the observed.
Types of Observation
Observation
varies in type and amount of information
to be collected.
It can be classified in relation to the settings in which data are collected, the structure of the observation, the researchers’ role, the method of data collection and so on
Open and Hidden Observation
Open observation
In open observation, the
identity of the researcher
, the recording equipment, and even the purpose of observation
are known to the observed.
Open observation is usually carried out in situations in which there is no need for, or no difficulty in obtaining people’s consent for, being observed.
Hidden observation
Hidden or covert observation is used when it is unlikely to gain access to open observation. It is difficult to obtain consent because the behaviour involved is considered deviant, illegal, or improper by many.
Also it is used when reactivity is likely to be a problem if the research is conducted openly. Observers and observation equipment, like camera or one-way mirror, are hidden from the scene.
Participant and Non-participant Observation
Participant observation
In participant observation,
researchers
enter into and are
involved in
the life or
environment of the observed
.
The level of participant observation determines the researcher’s degree of involvement in the observed target’s life.
Non-participant observation
In non-participant observation, researchers play the
role of pure observers.
They
do not have any interaction with the observed
.
They are complete strangers to them. They act as a completely objective and detached recording instrument observing and recording what happens.
Direct and Indirect Observation
Direct observation
In direct observation,
researchers observe (Specific) targets of people
, behaviour, or events that they intend to understand and explain.
For instance, if you want to know what people tend to do with the pamphlets they receive on the street, you observe people who have received the pamphlets.
Indirect observation
Indirect observation is used when
direct observation is not possible
or difficult, such as historical events that will not recur.
Researchers may
observe signs and physical details
in order to determine what has happened.
Natural and laboratory observation
Natural observation
This takes place in the natural setting in which the researched phenomena occur.
For instance, researchers who want to understand how people spend their time on buses actually go on buses, the natural setting, and carry out the observation.
Researchers may need to follow the observed targets and record what happens if the focus of study is on the targets’ behaviour in different situations.
The rationale for
keeping the natural setting is to maintain the observed target’s behaviour as close as possible to that when he or she is not observed
.
In this way, what is observed should be of
high validity
to what the researchers want to measure.
Laboratory observation
This is conducted in laboratories (or other similar confined environments) whereby factors conceived as
‘irrelevant’
to the research problem are controlled or
excluded
from the situations.
Researchers can then focus on specific variables and examine their relationships more accurately
Structured and Unstructured Observation
Structured observation
originates from the positive research traditions that
emphasize the precision
of definitions and operationalization of concepts,
accuracy and objectivity of data
collection, production of quantitative data and statistical analysis, etc.
It aims to produce accurate quantitative data on pre-set observable behaviours. The
data include frequency, duration, size, speed, etc.
Unstructured observation
aims to
produce detailed, qualitative
(and quantitative)
descriptions
of behaviour and settings.
It allows researchers to record data in a more flexible way.
It also enables researchers to explore the observed phenomena from different perspectives and attend to newly emerged categories and patterns of behaviour/events.
Recording Observational Data
The most popular method is keeping a
written record
. Observers may take notes to describe what they see and hear.
Video recording may also be used to record the process of events.
Tapes and videotapes
have several advantages over field notes: The recorded phenomena can be played as many times as required for coding data.
Chapter 9: Qualitative Research Design
What is Qualitative Research?
Qualitative research is research in which data are
mainly verbal description
and
conclusion does not rest on statistical analysis
.
It is a field of enquiry guided by certain epistemological viewpoint (i.e. subjective, interpretive approach).
Purpose of study Qualitative research is used
when the aim of study is
in-depth understanding
, and
when quantified data and
standardized procedures are not possible
Nature of Data
Data is verbal, rather than numeric. This is that data is expressed in words rather than in numbers. (Subjective as opposed to objective)
Verbal data, or words in research, are
based on observations, interviews, or documents.
Epistemology
Interpretive approach. (i.e. seek to understand values and meaning)
characteristics of qualitative research includes:
Holistic view
It is believed that complex phenomena
cannot be adequately understood by examining only some of their parts or a few factors
.
Rather, researchers have to gain a holistic (i.e. encompassing and integrated) overview of the phenomena.
open mind
Researchers should also free themselves from presuppositions and maintain an openness about what will be observed.
Subject oriented
The focus
should be on the perceptions of those being studied, not the researcher’s perceptions.
It is the perceptions of those being studied that are important.
We need not bother with the perspective of unrelated parties (e.g. the researcher) when trying to understand why a person does something in certain situations.
Conducted in the field
The researcher is supposed to carry out the investigation through an intense and/or prolonged contact with the ‘field’ or natural setting in which the phenomena occur, so that
essential information may be gathered in relation to its context and background,
holistic understanding enabled, and alteration of the phenomena minimized.
Setting
Natural setting
if researchers
want to understand meanings
that emerge from these interactions, they have to
enter into the field
or life situations of where the phenomena occur, or develop an adequate appreciation for the process for understanding the phenomena
Design
Research design in qualitative research is generally
less structured and more flexible
than research design in quantitative research.
Triangulation
The use of triangulation is not simply to include different kinds of data, but more importantly, to relate and
link data
in order
to counteract the threats to validity
identified in each
outlines various aspects in which triangulation may be used. They are:
Theory triangulation
— using
multiple
rather than single
perspectives
in relation to the studied phenomena.
Investigator triangulation
— the use of
multiple
rather the single
observers
Data triangulation
— the use of different sources of data instead of one such as collecting
data from different sources
(times and locations and from different people).
Methodological triangulation
— the use of
different data collection methods
Research Strategies of Qualitative Research
Ethnographic Research
Ethnographic research studies cultural groups
Ethnographic studies are characterised by:
using an empirical approach
keeping an openness of observation, and
studying the phenomena in situ/ in the field (i.e. in its original location).
Researchers frequently live with the people to be studied and try to become part of the culture.
A
good ethnographic study
should produce a final report that
enables readers to understand the culture
even though they do not have direct experience with it.
Issues related to Ethnographic Research: Resources, Ethics, Time
Ethics
The ethical issue is not simply whether the researcher maintains any secrecy in the role of the researcher, but how well the researcher performs his or her participant-observer role in a satisfactory way.
Generally, it is
difficult for field researchers to maintain a good balance
between the two demanding and rather non-complementary roles as the participant and the researcher.
(i.e. role as a researcher vs roled as a group participant)
Time
It is more difficult to conduct covert than overt ethnographic studies, partly because of the
time and energy spent on the social role, leaving little for the research
, and partly because of the occurrence of unexpected events that deter the research progress. Special skills and strategies are used to fulfill the roles of a researcher as well as a worker.
Despite the difficulties, covert participant observation offers researchers more thorough observation and in-depth understanding of meanings and feelings through empathetic understanding (understanding through experiencing similar events).
Resources
Tremendous resources are demanded
from qualitative researchers who investigate in the field.
Specifically, they require not only i
ntellectual knowledge
of how to collect or analyse data in specific ways, but also the
social awareness and skills required
for being part of the cultural group, self-discipline and motivation to maintain the research work as well as the social roles.
Grounded Theory Research
studies in which theories are developed in the process of collecting and analysing data (i.e. theories are grounded in, based on, concluded from, the data)
Assumption
presumes that
there are basic social processes
(i.e. fundamental patterns of all social life) and that it is possible to discover them through appropriate investigation.
Focus
The predominant focus of grounded theory is on an analytic approach to social science
Aim
the aim of the strategy is the
generation of theories
Main Steps in the Grounded Theory Research Process:
In jist it stats with data, to identifying concepts, to dcodifying concept, to interpretation
Assigning codes to concepts
: Codes are created and assigned to concepts as researchers closely examine the data.
Through coding, the data are defined and categorized
. The codes are vehicles for analysis and interpretation of data
Consulting the literature
: After concepts are identified and their relationships specified, researchers consult the literature to
see whether similar associations
have been found and discussed
Identifying pertinent concepts
: Concepts that are relevant to the research problems are identified and recorded.
Interpreting data:
Researchers keep an open mind and interpret the results by an intuitive process.
Plausible hypotheses are developed
, based on recurring concepts in the data.
Analysing data
early: Data are analysed
as soon as they are collected
.
Collecting data
: Data are recorded by means of
written notes and tape
recordings.
Terminating the process
: This process of data collection and analysis and refining the theoretical constructs continues until ‘saturation’ of data takes place.
Setting an area of study
: An area of study is set
without specific theories
or hypotheses when starting the research process.
Biographical Method
Focus
The biographical method
focuses on the study of an individual’s life
, and his or her perception of self and interpretation of social actions.
It relies on the use of
personal documents
such as diaries, autobiographies, personal journals, memoirs, and letters.
Characteristics
Biographical research may take a
holistic perspective
to analyse the document (i.e. forming overall impression of what the document tells) or particularistic perspective (i.e. focusing on specific aspects of the document).
The
analysis is basically qualitative
, whereby the manifest and latent (hidden) content of the document is systematically analysed to identify recurring patterns and main themes.
Biographical research
primarily deals with single cases
, but comparison of some cases may also be carried out to identify patterns of social phenomena
Documentary Method/ Meta-analysis
Characteristics
Documentary methods differ from other data collection methods in that the
researchers do not record the data they are going to analyse
.
Generally, the documents exist before the research is carried out, and the
researcher’s task is to select and get the appropriate existing dataset
to answer his or her research question.
Rationale
The rationale for the documentary method is that as long as the
existing documents contain the essential information
required for answering the research questions, they are useful data, no matter who collected them.
The selection and interpretation of existing datasets is processed under the assumption that the recorders of the
documents share some common meaning
with the researchers and do the recording in some standardized way, as the researchers do
Sources includes
Personal documents (e.g. life histories, diaries, autobiographies, and letters)
Formal research reports (e.g. journals, institutional research documents).
Official records (e.g. service records of doctors, social workers, counsellors, lawyers, and records of organizations)
Administrative documents (e.g. proposals, memoranda, progress reports, minutes of meetings, and announcements)
Public documents (e.g. census statistics, yearbooks, court archives, mass media, and literature such as novels and poetry)
Process of Qualitative Research
the qualitative research process is characterised by
Flexible sequence of steps:
The starting point (identification of phenomena to be studied) and the end point (final conclusion and report of results) is fixed, but
steps in between the sequence may be altered
a
hypothesis
may not be formulated at the early stage of the research (before data collection); rather, it may be generated and
modified throughout the data-collection process.
data collection and data analysis
usually take place
simultaneously
, which contrasts to what happens in quantitative research in which data analysis is pre-planned and usually takes place after the data are collected
On-going process of formulation of research problems
:
There is usually no clear-cut research question or hypothesis set before data are collected.
Hypotheses may be generated and grounded from the data.