Iuliana Lazar1*, Maria-Ema Faciu1, Liliana Mata1, Gabriel Lazar1

1Vasile Alecsandri University of Bacau (ROMANIA)

presents:

STRENGTHS AND LIMITATIONS OF

QUANTITATIVE RESEARCH APPLIED IN
THE EDUCATIONAL SCIENCES
Paper goals

Main aim:

Identifying the most common advantages and

limitations of quantitative research applied in the
educational sciences.
There are explored the main quantitative tools:

I. factory analysis,

II. clusters analysis

III. multiple regression technique

Present your project in Word,

PowerPoint, or Visio
Integrating the new technologies

The educational approach of integrating

the new technologies led to some
researchers preoccupation to create
innovative models based on correlating
technological and educational
components

Present your project in Word,

PowerPoint, or Visio
Integrating the new technologies

Mixed resources:

Interactive technologies (interactive whiteboard, online

educational platform, online collaborative applications) and
an online environment as a pedagogical component

and

a set of interactive and collaborative learning strategies

adapted to an online environment as a pedagogical
component
 Integrating the new technologies
Qualitative analysis was used to investigate the effects
produced by the use of the new technologies on school
results, students motivation or perception
To investigate teachers degree of acceptance or rejection
of mixed resources used in education (like modern learning
management system that is easy to use on ceramic
interactive whiteboards).

To provide to the persons involved in educational policies a

feedback related to implementation of educational
technologies in teaching-learning activities.

To test the efficacy of new training methods.

 Why Quantitative Research?

Goals

+
Here are some of the goals that mixed resources fulfills and can be
investigated with quantitative research:
Supporting the continuous learning process through a modern learning
management system that is easy to use both on ceramic interactive
whiteboards and educational platforms;
Providing support for the computer assisted instruction in various fields;
Creating a virtual space where the student and the teacher can meet;
Creating and presenting a set of interactive, appealing classes that fit the
needs of each student;
Easing the learning process by stimulating the students' creativity and their
competitive spirit;
Advantages

High accessibility regardless of the geographic area: the content is

available anywhere, at any time.
The individualization of the learning process.
Low maintenance costs.
The information is spread quickly.
The use of various multimedia learning resources and the internet
(short clips, games, simulations, photo galleries, videos, audio clips)
Paper contents
Context of the research
Research approach/question
Theoretical framework
Research methodology
Findings
Implications for teacher education
Context of the research
The need of this current study starts from
noticing the fact that quantitative research
have in the same time the strong points
and weaknesses. In order to performed a
good model to investigate an educational
research subject we must identify from
the start the proper tools and avoid
unnecessary ones.

Measurement:

Populations Size
Sampling
Random Assignment
Ethical rules
Context of the research
The need of this current study starts from
noticing the fact that quantitative research
have in the same time the strong points
and weaknesses. In order to performed a
good model to investigate an educational
research subject we must identify from
the start the proper tools and avoid
unnecessary ones.

Time:

Pilot study
Experts & repeated measures
Validate the model
Context of the research

In present, different kinds of multivariate analyses

are used to investigate variables behaviour that are
easy to identify and results can be generalized to
larger populations. In social sciences many
researchers prefer quantitative methods because
they provide clear objective based on data that can
be evaluated.
Context of the research
There have been done studies which approach the
development of some valid research tools based on
exploratory factor analysis or multiple regression models.
In the educational field, there can be identified studies
which use the principal component analysis, clusters
analysis and multiple regression for different
interpretations of the recorded experimental data.
However, interesting research topics, but who have low
results due to failed research design are still disseminated.
Context of the research

Methods for Quantitative research

Experiments and Quasi-experiments

Behavioral Measures

Questionnaires and Surveys

Context of the research

We present most relevant quantitative analyses

methods based on multivariate analysis which,
applied in the educational sciences, can complete
qualitative analysis to reveal latent information
based on results and can facilitate implementing
an integrated policy for education in development.

Also, this study provides a new approach on the

initial research design.

From this perspective, this theoretical study

presents the advantages and limitations of
quantitative analyses that we will have to take into
account for improving the research related to
perception of new technologies used in
education.
Defining quantitative research. EFA and CFA

Both exploratory factor analysis (EFA) and confirmatory

factor analysis (CFA) are used to understand common
variance of measured variables which are assigned to a
latent factor. However, EFA and CFA are conceptually and
statistically different.

EFA is a technique for identifying clusters of dependent

variables in order to understand the relationship between
the variables that we have measured. This type of
multivariate analysis allows the identification of latent
structure among the investigated variables. This thing is
done by estimating the variance common at their level.
Defining quantitative research. EFA and CFA

Each factor can be presented in terms of the variables

measured and the factor score regression coefficient.
More than that, the analysis of the factors is based on
Pearson linear correlation.

That is why the variables must present a linear

association, in order to be introduced in the analysis. In
many cases, we notice an abnormal distribution (the
negative asymmetry is predominant) for natural
dependent variables. But taking into consideration the
significant number of respondents, we will consider the
normality condition as not compulsory.
Defining quantitative research. EFA and CFA

The use of KMO index Kaiser-Meyer-Olkin represents a

measure of sampling adequacy and Barletts test of
sphericity, offers us information about the existence of
multicollinearity. There is a KMO index for each variable,
and their sum indicates the KMO global index.
Communality values are multiple correlation coefficients
for each variable as dependent variable, using the factors
as predictors.
Defining quantitative research. EFA and CFA

The higher value of KMO index as well as the significance

level of Barletts test of sphericity suggests the existence
of one or more common factors which justifies the use of
a factor reduction.
CFA allows testing the suitability of models established a
priori, for example, to verify the structure of the
instrument scale. Primary statistical analysis of the data is
used to test the factorial structure of the instrument.
Defining quantitative research. CA.

Cluster analysis, based on the results of factor analysis,

indicates that participants can be classified into several
clusters with particular characteristics in their behavior.
Thus, starting from the scores obtained from the
instruments, the respondents can be classified.
Defining quantitative research. CA.

The dependent variables are built on the basis of the

results of factor analysis. For each main factor, a global
score will be calculated by adding scores for each
included element and dividing it to the number of items
which represent the factor. Thus, all factors will have the
same scale and thus, standardization will not be
necessary anymore.
Defining quantitative research. CA.

The construction of clusters can be made by the method

of interactive updating of centers, either by partitioning
techniques, thus the distance among the internal
elements of groups should be smaller, and the distance
among groups should be bigger.

After the number of clusters was established, k means

clusters method is recommended, as being suitable for
medium size samples. The database of respondents can
be used as analyzed bodies and weighing factor analysis
as dependent variable.
Defining quantitative research. Regresion.

Regression models are used to identify the types and

degree of relation between the dependent and the
independent variables of the study. A regression model
can be established for each dependent variable
separately, based on a single or a combination of
independent variables (predictors).
Defining quantitative research. Regresion.

In order for the chosen model to explain the relationships

between variables, the value of determination ratio (R2)
should be between 0 and 1 and the value of statistical
significance index (Sig.) should be less than 0.05. When
the relationship between the dependent variable and the
independent variable is not linear, the estimation of the
regression curve is recommended.
STRENGTHS

the data are shown as descriptive terms and can be

used to create graphical models easier to interpret;
can be applied to large numbers of people;
allow the identification of latent structure among the
investigated variables;
STRENGTHS

contribuition of each variable as dependent variable can

be estimated;
the respondents can be classified under certain
categories (types) which present similar answer models;
LIMITATION

natural dependent variables usually presents an

abnormal distribution;
all variables with communality values lower than 0.5 are
recommended to be eliminated thus results can be lost;
LIMITATION

separate and randomized cases are needed to validate

and confirm the instrument scale, increasing the
investigation period;
quantitative research not explain why and how social
and economic factors contribute to education.
CONCLUSIONS

The purpose of factor analysis is to test a

hypothesis or theory about the structure obtained
on the basis of interdependencies among
variables.
This type of analysis enables the identification of
the main categories of factors, and groups the
participants.
CONCLUSIONS

Through the exploratory factor analysis,

data can be summarized by bringing
together the variables that were interrelated
in the early stages of research.

The validation of the measurement tool can

be performed by analyzing the internal
consistency with Cronbach alpha index.
CONCLUSIONS

Based on particular characteristic of the

participants, using cluster analysis they can
be classified into similar answers models,
while regression analysis can be used to
identify the types and degree of influence
of the particular characteristics.
Acknowledgements

This research was financially supported by the Executive Unit for

Financing Higher Education, Research, Development and
Innovation (Grant PN-II-PT-PCCA-2011-3.2- 1108,Networked
interactive ceramic whiteboards with integrated sound (ENO) for
teaching and learning science and technology).
Thank you for attention!