Unit XIX

Data analysis

Data analysis is the activity of transforming a set of data with the goal of being able to
verify them better by giving them at the same time a reason for being and a rational analysis. It is
analyze the data of a problem and identify them. Data analysis has different facets and
approaches, incorporating various techniques. It is of great importance in areas such as: sciences,
social studies and business, due to the diversity of possible models.

1.1 The importance of data analysis

All scientific work must seek to address a theme of scientific or academic relevance.
social and bring a new perspective on it, is what you will find if you search for articles on
Google Scholaror noScielo, for example.

In order for this objective to be met, there is amethodologywhich must be followed, for it is
I need to gather information and concepts on the subject in question, looking for materials on
reference and also conduct a critical analysis of everything that was found by the author of
work.

For this reason, it is essential that thedata collectionbe conducted with criteria and seriousness,
otherwise, incorrect conclusions may compromise all the work.

1.2 Methodology for data analysis

Analysis of quantitative research data

Forquantitative researchone can use statistical analyses, such as distributions of

frequency, correlations and graphical representations, measures of dispersion, measures of tendency
central, etc.

1.3 Analysis of qualitative research data

Already in the case of thequalitative

researcha content analysis can be made that will serve as
basis for a qualitative analysis of the interviews, observations, and open questions in
questionnaires, etc.

This analysis seeks to understand the meaning of the collected data and also aims to
facilitate the understanding of the contents through some classification presented in a way
systematized, which can be the division into categories or by the counting of words and terms
contained in the answers.
1.4 Types of data analysis

1.4.1 Predictive Analysis

This is the most well-known model, as it helps to predict future scenarios based on the analysis of
database standards. Thus, it is possible to make more precise decisions.

The methods used by predictive analysis are statistical and historical data, as well as data mining.
of data and artificial intelligence.

Thus, it is indicated to project future behaviors of the audience and the market, in addition to
evaluate the fluctuations of the economy and consumption trends.

1.4.2 Prescriptive analysis

The consequences of the actions taken are verified, which allows knowing what should occur to
choose certain attitudes.

Therefore, this layer is the one that holds the most value, as it requires the human element to be
concretize.

In addition, it is relevant because it defines the path to be taken for the action to occur.
as expected.

In other words, a goal is set and from that, the paths that should be taken are indicated.
to be traveled to reach it.

Therefore, prescriptive analysis is considered the most complex, as the professional must be knowledgeable.
data science techniques and being an expert in the business and the ecosystem in which it is inserted.

1.4.3 Descriptive analysis

The objective of this model is to allow for the understanding of events in real time. It is very
used in situations such as credit analysis.

In this case, the bank evaluates the individual's information and assesses the risk involved in the process.
Thus, the interest rate is defined.

As descriptive analysis does not make value judgments, it is indicated for visualizing data.
and understand the impact in the present, but without making a connection with the past or the future.

Moreover, it helps to make immediate decisions with calmness and security.

1.4.5 Diagnostic analysis

Aiming to understand the causes of an event, you might set out to answer questions such as:

Who?
When?
Where?
How?
Why?

The ideal is to analyze the impact and reach of an action taken. From this, one can outline
strategies to improve results. This is a model widely used in sales and should be
supplemented with predictive analysis to reinforce data projection.

How are data analyses performed?

To carry out these types of data analyses, we need to go through some processes, among which are
which:

1.4.5.1 Exploratory analyses

The analysesexploratoryare made when the data has not yet been integrated or may be
incomplete. They are not fully automated because they need someone to verify.
outliers for these data to be inserted into systems.

Data modeling

During the modeling selection, automation resources are already used. It is here that the
professionals determine what the best approach is, that is, what the ideal type of analysis for
a task.

Moving on to the modeling itself, the rules for predictive and descriptive analyses are created.
diagnostic and prescriptive.

1.4.5.2 Report generation

Umsummarytwo data, after the type of analysis performed, must be generated to guide the
business decisions. This summary takes into account the clarity of the data used and the
accuracy of the analysis performed.

All types of data analysis have a role within organizations. Each of them meets
to a specific need, guides managers in a direction and is applied to a certain
context.
The assistance of a data scientist is essential to establish the best type of analysis in
Each case and the hiring of this professional is strategic for businesses to achieve more with
the information they have.

1.4.5.3 Exploratory research

By investigating the data and processes with an open mind, you can generate hypotheses.
about the causes.

Using, for example, the new knowledge found to identify the main causes
likely for the defects.

In addition to testing hypotheses or eliminating causes using data, experiments, or more

analysis of the process to verify which potential causes contribute significantly to
the problem.
2. Content Analysis

Content Analysis is a tool that allows the researcher to study behavior

human indirectly, through the analysis of their communications. Usually, they are analyzed
the written contents of a communication, but, for example, an image or a sound can
be the focus of a content analysis (Fraenkel & Wallen, 2008). Periodicals, articles, films,
music, graffiti, photos, handicraft objects, in short, a series of communication types
that reflect human behavior can be the subject of a content analysis.

Generically, Denscombe (1998) characterizes this instrument as a resource that helps the
investigator analyzing the content of documents, applicable to any content
of communication, reproduced through writing, sound, or image.

Currently, with the evolution of new technologies, namely web tools

2.0, communication sources are increasingly diverse. Blogs (posts and comments), wikis,
online communities and 3D virtual environments, for example, are increasingly becoming the target of
investigation and, consequently, its contents subject to analysis (Gray, 2004).

To conduct a Content Analysis, the researcher needs to organize a sample.

considerable amount of material. But how to do this? Fraenkel & Wallen (2008) indicate that it is through
the development of a categories system that the researcher can use for later
comparison in order to illuminate what is being investigated.

According to Berelson (1968), cited by Carmo and Ferreira (1998), Content Analysis "allows
make an objective, systematic, and quantitative description of the manifest content of
communications, aiming for their interpretation." Thus, the description of the content is
objective in the sense of the effort of analysis to follow clear rules and instructions that allow for
reproduction of the investigation among the investigators, that is, it makes it possible to reach the same ones.
results when working on the same content. The description of the content is also systematic,
since the content is organized and integrated into a categorization system according to the
research objectives. Finally, the description is quantitative, as in the categories
generally, the frequency of the elements considered relevant for the
investigation.

Bardin (2004) deepens the knowledge about Content Analysis and emphasizes that this should
go beyond the mere description of the content of messages, and include the inference of knowledge
about the conditions of production/reception of content with the support of indicators. Thus, the
the analysis process primarily involves an effort of description, where the characteristics of
communication is worked on, followed by an inferential effort, which allows one to move from the
description for interpretation, that is, attribution of meaning to these characteristics.

Content Analysis, as an interpretative effort, seeks to balance rigor

objectivity and the richness of subjectivity. According to Bardin (2004: 37), this type of analysis is:
a set of communication analysis techniques aimed at obtaining, through procedures
systematic and objective content description of messages, indicators (quantitative or
not) that allow the inference of knowledge related to reproduction/reception conditions
(inferred variables) of these messages." Inferences can be made about both the sender
as for the communication receiver, that is, about the origin of the message and of itself
recipient.

Content Analysis considers the articulation between the description and analysis of the described text, and the
logical deduction of the factors that determined the characteristics of the characteristic elements.

2.1 Phases of Content Analysis

Focused on the objectives of the study and the theoretical framework, Content Analysis is
carried out through a series of steps. Based on the authors Bardin (2004), Carmo & Ferreira
(1998) and Pardal & Correia (1995), the following phases are established:

Definition of categories to separate the observable data;

Definition of units of analysis;

Distribution of the units of analysis by the previously established categories;

Interpretation of the results obtained from qualitative and/or quantitative perspectives.

2.2 Types of Content Analysis

Grawitz (1993) distinguishes three groups of Content Analysis:

1. Exploration analysis and verification analysis

Here, two different purposes of analysis confront each other: - the verification of a hypothesis,
where the objective is clearly defined and results in the quantification of outcomes; - that of exploration,
where there are no predefined hypotheses, and it allows for various outcomes. However,
the exaggerated systematization of analysis can leave out elements from the field of study
essentials that were not predicted in advance.
2. Quantitative analysis and qualitative analysis

The quantitative focuses on the frequency of the characterized elements, while the qualitative focuses
about the value of a theme, the novelty, the interest.

3.Direct analysis and Indirect analysis

Related to quantitative analysis, the direct method involves a simpler procedure.

where one generally resorts to the comparison of frequencies (number of occurrences) of certain
elements under analysis. Indirect analysis is more related to the qualitative nature, and
Are you interested in an interpretation of what lies behind expressed language.

2.3 Relevant Aspects of the Analysis

According to Fraenkel & Wallen (2008), Content Analysis can be used, for example, when
try to describe trends in schooling; understand organizational patterns (by
example, through the analysis of documents prepared by the administrators); demonstrate how
various schools face the same phenomenon, but in different ways; to understand what the
teachers feel about their profession.

Wilkinson & Birmingham (2003) classify content analysis as a tool for

investigation that can be applied when the investigator is confronted with the
results of an investigation or with your own. For example, what to do with the data
collected during the interview? What to do with the researcher's diary notes? What to do with
the images produced through the Photovoice technique? These are just some questions that the
researchers encounter when applying their techniques and data collection instruments.

2.4 Advantages

Help answer questions like: what do we do with the data collected from the interview?
What to do with the investigator's diary notes? Content analysis presents itself as a
an extremely useful technique for analyzing data from interviews and observations. According to Fraenkel
According to Wallen (2008), the main advantage of content analysis is the absence of intrusion.
As the investigator interacts with materials (documents, sound, image), he/she can 'observe'
without being observed since there is no influence of the investigator's presence here. The information that
it can be difficult, or even impossible to obtain through observation, it can be analyzed without
that its author is aware that he is being analyzed.

Another positive aspect is that it allows the researcher not to be tied to time and space to
the study of present events. The investigator can retrieve/alter records to obtain some
meaning for social life in a more current time.
The logistical aspects are also advantages of this technique as it is generally simple and
economic. This is well characterized by the treatment of newspapers, magazines, and books, for example.
Another potentiality is that this technique allows for the replication of the same research by another.
researcher, once the data is available and can almost always be "returned".

2.5 Limitations

There is usually a tendency among researchers to consider that interpretations of

a given content analysis indicates that it is more a cause of a phenomenon than a reflection
about this (Bardin, 2004).

Related Instruments

Other instruments that are often presented in conjunction with Content Analysis:

Investigator's Diary
Interview

Questionnaire

Checklists

Reading Cards

Bibliography

Bardin (2004). Content Analysis. 3rd Ed. Lisbon: Edições 70.

Carmo, H.; Ferreira, M. (1998). Research Methodology: A Guide for Self-Learning.

Lisbon: Open University.

Denscombe, M. (1998). The Good Research Guide for small-scale social research projects.
Philadelphia: Open University press.

Fraenkel, J.; Wallen, N. (2008). How to Design and Evaluate Research in Education. 7th Ed.
New York: McGraw-Hill International Edition.

Doing Research in the Real World

Pardal, L.; Correia, E. (1995). Methods and Techniques of Social Research. Porto: Areal Publishers.
Wilkinson, D.; Birmingham, P. (2003). Using Research Instruments: A Guide for Researchers.
London: RoutledgeFalmer.