Handouts in Research 2
Handouts in Research 2
Handouts in Research 2
1
RESEARCH IN DAILY LIFE 2
Quantitative experiments are useful for testing external validity and consider these factors in their
the results gained by a series of qualitative interpretation.
experiments, leading to a final answer and a a. Non-Equivalent Control Group Design
narrowing down of possible directions to follow. - refers to the chance failure of random
assignment to equalize the conditions by converting a
WEAKNESSES OF QUANTITATIVE RESEARCH true experiment into this kind of design, for purposes of
Quantitative research can be costly, difficult analysis.
and time-consuming. Difficult because most
researchers are non-mathematicians. b. Time Series Design
Quantitative studies require extensive - employs multiple measures before and after the
statistical treatment requiring stringent experimental intervention. It differs from the single –
standards, more so with confirmation of results. group pre-experiment that has only one pre-test and
Quantitative methods also tend to turn out only one post-test. Users of this design assume that the
proved or unproven results, leaving little room time threats such as history or maturation appear as
for uncertainty or grey areas. regular changes in the measures prior to the
intervention.
KINDS OF QUANTITATIVE RESEARCH DESIGNS
Research design refers to the overall strategy 3. Pre-Experimental Design
that you choose in order to integrate the different a. One-Shot Case Study or One Group Posttest
components of the study in a coherent and logical way Only Design
thereby ensuring you will effectively address the - In this design, a single group of individuals is
research problem. measured on some dependent variable after an
Quantitative methods emphasize objective intervention has taken place. The researcher has to
measurements and the statistical, mathematical, or evaluate the experiment by interviewing people (O)
numerical analysis of data collected through polls, and trying to assess the impact of the intervention (X).
questionnaires, and surveys, or by manipulating pre-
existing statistical data using computational b. One Group Pretest – Posttest Design
techniques. - The one-group pretest–posttest design usually
involves three steps: (1) administering a pretest
QUANTITATIVE RESEARCH DESIGNS measuring the dependent variable; (2) applying the
A. EXPERIMENTAL DESIGNS experimental treatment X to the subjects; and (3)
Experimental research design allows the administering a posttest, again measuring the
researcher to control the situation. In so doing, it allows dependent variable. Differences attributed to
the researcher to answer the question “WHAT application of the experimental treatment are then
CAUSES SOMETHING TO OCCUR?” This kind of evaluated by comparing the pretest and posttest
research also allows the researcher to identify the scores.
cause and effect relationship between variables and to
distinguish the placebo effects from treatment effects. B. NON-EXPERIMENTAL DESIGNS
Descriptive Research Designs
1. True Experimental Design In this kind of design, the researcher observes
True experimental designs (also called the phenomena as they occur naturally and no
randomized designs) use randomization and provide external variables are introduced. In this research
maximum control of extraneous variables. design, the variables are not deliberately manipulated
a. Pre-Test Design nor is the setting controlled.
b. Post Test Design The Descriptive Research Design’s main
c. Post Test Only / Control Group Design purpose is to observe, describe, and document
aspects of a situation as it naturally occurs and
2. Quasi-Experimental Design sometimes to serve as starting point for hypothesis
Quasi-experimental designs are similar to generation or theory development.
randomized experimental designs in that they involve
manipulation of an independent variable but differ in Types of Descriptive Research Designs
that subjects are not randomly assigned to treatment a. Survey
groups. Because the quasi-experimental design does - A research design used when the researcher
not provide full control, it is extremely important that intends to provide quantitative or numeric description
researchers be aware of the threats to both internal and of trends, attitudes, or opinions of a population by
studying a sample of that population (Cresswell, 2003).
2
RESEARCH IN DAILY LIFE 2
b. Correlational goal of developing a scale – matched approach, where
Correlational research has three types: data from different disciplines can be integrated.
Bivariate Correlational Studies – obtain scores
from two variables for each subject then use them to IMPORTANCE OF QUANTITATIVE RESEARCH
calculate a correlation coefficient. The term bivariate ACROSS FIELDS
implies that two variables are correlated (variables are
selected because they are believed to be related). Quantitative Research across Disciplines
Quantitative Research and Anthropology
Examples: Children of wealthier (V1), better educated Quantitative Research and Communication
(V2) parents earn higher salaries as adults. Quantitative Research and Sports Medicine
Quantitative Research and Medical Education
Prediction Studies – use correlation co-efficient Quantitative Research and the Behavioral
to show how one variable (the predictor variable) Sciences
predicts another (the criterion variable). Quantitative Research and Education and
Psychology
Examples: Which high school applicants should be Quantitative Research and the Social Sciences
admitted to college?
KINDS OF VARIABLES AND THEIR USES
Multiple Regression Predictor Studies – Bernard (1894) defines a variable as something
suppose the high school GPA is not the sole predictor that can take more than one value and value can be
of college GPA, what might be other good predictors? words or numbers.
All of these variables can contribute to the overall A variable specifically refers to a characteristic or
prediction in an equation that adds together the attribute of an individual or an organization that can be
predictive power of each identified variable. measured or observed and that varies among the
people or organization being studied. (Creswell, 2002).
c. Ex Post Facto Studies
- These are non-experimental designs that are The Nature of Variables and Data
used to investigate causal relationships. They examine Generally speaking, variables can be classified
whether one or more pre-existing conditions could as one of the four types:
possibly have caused subsequent differences in 1) Nominal variables represent categories that
groups of subjects. Researchers attempt to discover cannot be ordered in any particular way.
whether differences between groups have results in an 2) Ordinal variables represent categories that
observed difference in the independent variables. can be ordered from greatest to smallest.
3) Interval variables have values that lie along an
d. Comparative Design evenly dispersed range of numbers.
- involves comparing and contrasting two or 4) Ratio variables have values that lie along an
more samples of study on one or more variables, often evenly dispersed range of numbers when there is an
at a single point of time, Specifically, this design is used absolute zero as opposed to net worth, which can have
to compare two distinct groups on the basis of selected a negative debt-to-income ratio level variable.
attributes such as knowledge level, perceptions, and
attitudes, physical, or psychological symptoms. For Kinds of Variables
example a comparative study on the health problem Several experts have lumped together that
among rural and urban older people from district following as the major kinds of variables.
Mehsana, Gujarat. 1. Independent Variables – those that probably
cause, influence, or affect outcomes. They are
e. Evaluative Research invariably called treatment, manipulated, antecedent,
- seeks to assess or judge in some way of or predictor variables.
providing information about something other than 2. Dependent Variables – those that depend on
might be gleaned in mere observation or investigation the independent variables; they are the outcomes or
of relationships. For example, where a test of children results of the influence of the independent variable.
in school is used to assess the effectiveness of 3. Intervening or Mediating Variable – “stand
teaching or the deployment of a curriculum. between” the independent and dependent variable and
they show the effects of the independent variable on
f. Methodological the dependent variable.
– in this approach, the implementation of a variety 4. Control Variables – special types of
of methodologies form a critical part of achieving the independent variable that are measured in a study
3
RESEARCH IN DAILY LIFE 2
because they potentially influence the dependent 3) Apply to specific context.
variable. 4) Define the aim or desired outcome of this study.
5. Confounding Variables – those that are not
actually measured or observed in a study. They exist Choosing a Topic for a Research Study
but their influence cannot be directly detected in a Layder (2013) stresses that in choosing a
study. research topic, it is important to select a topic that
interests and motivates you. It should It should be
Chapter 2 “doable” with the resources at your disposal and within
IDENTIFYING THE INQUIRY AND STATING THE the required time frame. Needless to say, the research
PROBLEM topic should be focused and well-defined.
4
RESEARCH IN DAILY LIFE 2