Descriptive-Comparative
research designs are appropriate
for collecting descriptive information about a population of subjects of a
study. They are appropriate for descriptive studies, like profile studies,
exploratory studies, and doing small case studies. They are also ideal for
diagnostic studies or situation analysis. However, they are not recommended for
evaluation studies intended to determine the effect or impact of a certain
intervention or treatment.
The test of difference is a statistical measure that aims to determine
if there is a significant difference among variables. In the test of the difference, you are comparing one variable to the other, like comparing which
schools are performing well in terms of licensure examinations, comparing which
department has the high job satisfaction rate, and etc.
The statistical measure depends
on the nature of the sample and hypothesis testing procedure. The nature of the sample depends on whether the
samples are dependent or independent or whether the samples are only two or
more than two samples. The nature of the sample involves whether the samples
are independent or different samples
being compared to (i.e. male or female; school A, school B, or school C, and
section 1 or section 2) and dependent
or the same samples being compared at (i.e. grades of the students before and
after a certain intervention or awareness of the students before and after the
awareness seminar). The nature of the sample also depends on the number of samples, whether there are two or more
than 2 samples.
Moreover, the statistical
measure depends on whether the hypothesis testing is a parametric or
non-parametric test. A parametric test is
a hypothesis testing procedure based on the fulfillment of certain assumptions,
like normality, linearity, homogeneity, etc. On the other hand, it does not
require fulfilling the mentioned assumptions.
The application of statistical measures for the test of
difference depends on the following:
1.
Nature
of the sample
1.1.
Independent or dependent sample
1.2.
2 samples or more than 2 samples
2.
Nature
of hypothesis testing procedure
2.1 Parametric or
Non-parametric test
Below are some statistical tests for exploring differences
between groups:
Example 1
Research Title:
Peer Mentoring Program on Academic Performance of the First Year Students of St. Francis College: A Comparative Study
Statement of the Problem:
1. What is the level of the academic performance of the respondents?
2. Is there a significant difference in academic performance before and after the peer mentoring program was implemented?
Hypothesis:
There is no significant difference in academic performance before and after the peer mentoring program was implemented.
Conceptual Paradigm:
Research
Design: Descriptive - Comparative
Statistical
Analysis: T-test for Dependent Samples for
Parametric Test*; Wilcoxon Signed Rank test for Non-parametric
test
Instrument: Academic
performance of the respondents before and after the peer mentoring program was
implemented.
Example 2
Research
Title: Competence of Radiologic Technologists Employed in Private and
Public Hospitals
Statement of the Problem:
1. What
is the demographic profile of the respondents when grouped according to:
1.1 Age
1.2 Number of
years in service
1.3 Marital status
2. What
is the level of competence among the respondents employed in private and public
hospitals in terms of:
2.1 Knowledge
2.2 Skills
2.3 Attitude
3. Is
there a significant difference in the level of competence among respondents
when grouped according to:
3.1
Private hospital
3.2
Public hospital
4. Does
the demographic profile of the respondents significantly moderates when grouped
according to:
4.1 Age
4.2 Number of
years in service
4.3 Marital status
Hypothesis:
1. There
is no significant difference in the level of competence among respondents when
grouped according to:
1.1
Private hospital
1.2
Public hospital
2. The demographic
profile of the respondents does not significantly moderate when grouped
according to:
2.1 Age
2.2 Number of
years in service
2.3 Marital status
Conceptual Paradigm:
Research Design: Descriptive - Comparative
Statistical Analysis: T-test
for Independent Samples for Parametric Test*; Mann-Whitney U test for
Non-parametric test
Instrument: Questionnaire
to determine the level of competence of the respondents
Example 3
Research Title: Physical Therapy Licensure Examination Results Among Tertiary Schools in Davao City: A Comparative Study
Statement of the Problem:
1. What is the level of licensure examination results
in Physical Therapy among tertiary schools in Davao City?
2. Is there a significant difference in the level of
licensure examination results in Physical Therapy among tertiary schools in
Davao City?
Hypothesis:
There is no significant difference in the level of licensure
examination results in Physical Therapy among tertiary schools in Davao City.
Conceptual Paradigm:
Research Design: Descriptive - Comparative
Statistical Analysis: Analysis of Variance (ANOVA) for Parametric
Test*; Kruskal Wallis for
Nonparametric test
Instrument: Licensure examination
result
*The following assumptions should be fulfilled first before using Parametric Test:
1. Homogeneity of variance.
2. The sampling distribution of the means should be normally distributed.
