INVESTIGATION METHODOLOGY

ANALYSIS OF DATA

I. WHAT DOES IT CONSIST OF?

Data analysis consists of carrying out the operations to which the Researcher will subject the
data in order to achieve the objectives of the study.

All these operations cannot be rigidly defined in advance. Data collection and certain preliminary
analyzes may reveal problems and difficulties that will make the initial data analysis planning out
of date.

 However, it is important to plan the main aspects of the analysis plan based on the verification
of each of the formulated hypotheses since these definitions will in turn condition the data
collection phase.1

II. METHODOLOGY IN DATA ANALYSIS

Carrying out statistical studies involves issuing quantifiable results of said study or
experiment.

The clarity of this presentation is of vital importance for the understanding of the results and
their interpretation.

When representing the results of a statistical analysis in an appropriate way, numerical data
will be presented through tables; sometimes a diagram or graph can help us represent our data
more efficiently.

III. DATA ANALYSIS TECHNIQUES

There are two large families of data analysis techniques:

 Qualitative techniques: in which the data are presented verbally (or graphically) such
as interview texts, notes, documents.
There are no formal rules (in the style of statistical methods) for conducting
qualitative analyses. However, these studies are usually carried out in the following
four stages:

a. Preparation and description of raw material

b. Data reduction
c. Choice and application of analysis methods
d. Cross-sectional analysis of the cases studied (if there is more than
one)

1
http://www.ccee.edu.uy/ensenian/catmetinvcont/material/PPT%20Parte%207%20Analisis%20de%20los
%20datos-1.pdf
  Quantitative techniques: in which data is presented in numerical form.

IV. ANALYSIS OF DATA IN SURVEYS

The analysis of survey data, like any other type of data of scientific interest, must be
related to the knowledge problem that is being clarified and to the metric of the
empirical information that is at hand, that is, the The first thing to do in a survey is not
to see what the data says but what it says in relation to the problem that arises and
the hypotheses that one had previously posed.

V. TYPES OF DATA ANALYSIS IN SURVEYS

The analysis of survey data aims to detect highly related variable groups, for which the following
analyzes are used:
 Descriptive Analysis:
It will help to observe the behavior of the sample under study, through tables, graphs
.
The results collected in the sample are summarized in an N x M data matrix, in which
N is the number of units of analysis used (number of cases) and M is the number of
characteristics of said units, units of which we have information.
 Exploratory Analysis:
It aims to start from a deep and growing knowledge of the data to, working
inductively, arrive at an adjusted model of the data.

 Confirmatory/Explanatory Analysis:
Most of the traditional statistical data analysis techniques have a confirmatory
deductive nature. Of all the confirmatory analysis techniques, the most useful for the
sociologist is that which begins with the analysis of variables, among which it is worth
distinguishing: nominal and interval or ratio. In sociological research, ordinal
variables are very frequent, but analysis techniques are scarce, so they are treated
as interval or ratio variables, a very complicated task, or as nominal.

VI. DATA ANALYSIS PROCEDURES

Once the data collection and processing stages are completed, one of the most important
phases of an investigation begins: data analysis. At this stage it is determined how to analyze
the data and what statistical analysis tools are appropriate for this purpose. The type of data
analysis depends on at least the following factors.

a. The level of measurement of the variables.

b. The type of hypothesis formulated.

c. The research design used indicates the type of analysis required to test hypotheses.

Data analysis is the precedent for the interpretation activity. The interpretation is carried out in
terms of the research results. This activity consists of establishing inferences about the
relationships between the variables studied to draw conclusions and recommendations
(Kerlinger, 1982). The interpretation is carried out in two stages:

a. Interpretation of the relationships between the variables and the data that support them
based on some level of statistical significance.

b. Establish a broader meaning of the research, that is, determine the degree of
generalization of the research results.

The previous two stages are based on the degree of validity and reliability of the research. This
implies the ability to generalize the results obtained.

“Analyzing means establishing categories, ordering, manipulating and summarizing data,”

(Kerlinger, 1982, p. 96). In this stage of the research process, the data collected is rationalized
in order to explain and interpret the possible relationships expressed by the variables studied.

The design of statistical tables allows complex analysis techniques to be applied, facilitating this
process. The analysis should be expressed clearly and simply using both inductive and
deductive logic.

The results of research based on sample data require an approximation to the true value of the
population (Zorrilla, 1994). To achieve the above, a series of statistical techniques is required.
These techniques are derived from both parametric statistics and non-parametric statistics. The
first assumes that the population studied has a normal distribution and that the data obtained
were measured on an interval and ratio scale. The second does not establish assumptions
about the distribution of the population, however, it requires that the variables studied be
measured at a nominal or ordinal level (see Weiers, 1993).

Tables designed for data analysis are included in the final report and can be useful to analyze
one or more variables. By virtue of this last criterion, data analysis can be univariate, bivariate or
trivariate depending on the number of variables that are analyzed.
VII. STATISTICAL ANALYSIS WITH SPSS

o The SPSS statistical package makes it easy to obtain information through a survey,
summarize that information, and perform analysis and interpretation of the data.

o It is essential that every researcher handle this package or similar packages to achieve
higher levels of competence, as well as support their conclusions and assertions in any
field of research.

o Through SPPS you can:

• Design the phases of a survey, and questionnaire design.

• Prepare data analysis.
• Build the data matrix.
• Data purification.
• Error evaluation.
• Creation of data files.