Best data collection methods for research
The best data collection methods for research are very important. An investigation is scientifically valid when it is supported by verifiable information, which answers what is intended to be demonstrated with the formulated hypothesis. To do this, it is
essential to carry out a data collection process in a planned manner and with clear objectives regarding the level and depth of the information to be collected. This article presents
a series of criteria to consider when designing the information collection tool, as well as the collection methods to achieve
reliable results in an investigation.
Statistical units or individuals are called the component elements of the population studied. The population can be a group of human persons: the staff of a
company, the students of a teaching center, the Inhabitants of Spain on December 26, 1996 or a group of objects, the production of a workshop, the Spanish automobile fleet set of invoices of a company, etc. The population in a statistical investigation must be defined precisely.
Each of the individuals in the population can be described according to one or more characters. Thus, in the case of students from a center we could study the characteristics:
sex, age, grade, pending subjects, previous half year, number of siblings, address. The characters can be numeric in which case we will call them variables (quantitative) or non-numeric attributes (qualitative).
Each of the characters studied can present two or more modalities. The modalities are the different possible situations of the character. The modalities of a
character must be at the same time incompatible and exhaustive. That is to say, each individual in the population can adopt one and only one modality of character.
The simplest characters admit two modalities (dichotomous characters), for example sex (Male, Woman), a piece satisfies the quality requirements or not. The
number of modalities of a character can vary depending on the information that is to be collected, for example marital status:
• With two modalities: Single or Married
• With three modalities: Single, married or divorced
• With four modalities: Single, married or divorced, widowed •
With five modalities: Single, married or divorced, widowed, undeclared.
There are some characters that, due to their nature, the number of modalities is very large, for example the profession . It is difficult to determine the number of modalities that this
character has, so it is convenient to group them into category systems, so that each profession is framed unequivocally in one and only one of the aforementioned
categories. Thus in Spain they are divided into four main categories:
a) Agriculture
b) Construction
c) Industry
d) Services
These categories are in turn subdivided into other subcategories, so that any profession is included within a single category and within this in a single subcategory. The Category System must be exhaustive and mutually exclusive, to prevent an individual from belonging to more than one category.
Information Sources: These are all those media from which information comes, that satisfy the needs for knowledge of a situation or problem presented,
which will subsequently be used to achieve the expected objectives.
According to their origin they are classified into:
Primary sources.
Secondary sources.
Primary sources are those in which the data come directly from the population or sample of the population, while secondary sources are those that are
based on pre-prepared data, such as data obtained from statistical yearbooks, the Internet, media. Communication.
In turn, primary sources can be subdivided into:
1. Direct observation.
2. Indirect observation.
Direct observation is when the researcher directly takes data from the population, without the need for questionnaires or interviewers. For example, when a teacher
carries out a statistical study on the performance of her students.
Observation is indirect when the data are not obtained directly by the researcher, since it requires a questionnaire, interviewer or other means to obtain
the study data. For which it is necessary to carry out a survey.
The Primary Sources for compilation are obtained through direct research into the object of study, through established methods. To gather primary data,
the ideal is to use a plan that requires making several decisions:
- Research methods and instruments.
- The sampling plan
- Techniques to establish contact with the
public.
Secondary Sources to be used must be analyzed under 4 basic questions, which are:
• Is it relevant? when the information adapts to the objectives
• Is it obsolete? when it has lost relevance
• Is it reliable when the veracity of the source of origin is not questioned
• and is it trustworthy? if the information has been obtained with the
appropriate methodology and necessary honesty, with objectivity, continuous nature and accuracy
The data collection phase is considered to be the highest cost stage, both in human and material resources, as well as time. It is the most susceptible to
error.
DATA COLLECTION METHODS
To collect primary data in scientific research, we basically proceed by observation, by surveys or interviews with the study subjects, and by experimentation.
Survey:
It constitutes the middle ground between observation and experimentation. In it, situations that can be observed can be recorded and in the absence of being able to recreate an
experiment, the participant is questioned about it. Therefore, it is said that the survey is a descriptive method with which ideas, needs,
preferences, usage habits, etc. can be detected.
The survey is defined by Professor García Fernando as “an investigation carried out on a sample of subjects representative of a broader group, using
standardized interrogation procedures in order to obtain quantitative measurements of a wide variety of objective and subjective characteristics of the population.” .
Applying a survey to a representative sample of the population is with the aim of obtaining results that can later be transferred to the entire population.
Among the fundamental characteristics of a survey are:
1.- The survey is a non-direct observation of the facts through what the interested parties say.
2.- It is a method prepared for research.
3.- It allows a massive application that, through a sampling system, can be extended to an entire nation.
4.- It makes it possible for social research to reach the subjective aspects of the members of society.
DESIGN OF A QUESTIONNAIRE
The questionnaire is a set of questions about the facts or aspects that are of interest in an investigation and that are answered by the respondents. It is a
fundamental instrument for obtaining data. The objective of a survey design is to minimize the non-sampling errors that may occur, in that sense.
The questionnaires must have the following characteristics:
1. Operational: Easy to handle, used as a data collection instrument, providing in its structure the ease of emptying or outputting the information. Easy to process and tabulate.
2. Reliable: That it is reliable, that it allows the real collection of the objectives and that they
are easy to be debugged.
3. Valid: It must be concise, clear, firm, consistent, and not open to ambiguity.
Clear, brief, concrete and logical questions.
To design a questionnaire, three aspects must be considered:
1. The type of Questions and the order in which they should be grouped.
2. Formulating the questions according to the objectives, writing them grammatically.
3. The organization of the questionnaire material, putting yourself in the place of the interviewee.
Types of Questions in a Questionnaire
1. Dichotomous: It is the simplest and is used as a filter. Only admit as an answer: Yes or No.
2. Multiple Choice: Allows you to choose several answers within a series of answers.
3. Open: Leave the interviewee free to respond as he or she considers appropriate.
4. Closed: In this the interviewee can only choose one answer from a series of answers.
Pilot Questionnaire: It is one that, once structured and written, is subjected to the necessary tests in field work, or its simulation, with the main objective of
determining its validity as a measurement instrument and its degree of adaptation to the environment or universe investigated. .
Once the test has been carried out, the necessary corrections or changes will be made so that it becomes the ideal instrument for data collection.
The parts that a questionnaire must include are:
to. Greeting (which includes the reason for the questionnaire and the confidentiality with which the results obtained will be worked on)
b. Title
c. Instructions to answer it
d. Gratitude
To make a good questionnaire, experience plays an important role since creating a good questionnaire is considered an “art”.
It usually comes in two forms.
- Individual questionnaire: It is in which the respondent answers individually and without the interviewer directly intervening.
2. List Questionnaire: The questionnaire is asked of the respondent by a research specialist, who writes down the answers on a sheet with a series of rows and columns in which the different answers of each of the interviewees are noted.
THE INTERVIEW
The interview is the most important research instrument, along with the construction of the questionnaire. In an interview, in addition to obtaining the subjective results of the respondent regarding the questions in the questionnaire, the surrounding reality can be observed, with the interviewer noting down, in addition to the answers as they come out of the mouth of the interviewee, the aspects that he considers appropriate throughout the interview. the interview.
The interview, unlike the questionnaire that is answered in writing by the respondent, has the particularity of being more specific, since the questions presented in a forceful manner by the interviewer
leave no ambiguities, it is personal and not anonymous, it is direct because it is not Let the respondent consult the answers.
It is a convenient method to obtain data regarding the population, provided by individuals and that helps us understand social reality. These data could be observed
directly through observation but would be subjective to the researchers, making them more expensive to obtain. Interviews can be classified according to their degree of standardization into interviews with a questionnaire and without a questionnaire.
When you want to define the first information about the object of the survey, an unstructured interview, without a questionnaire, is convenient, which allows the researcher
to define the problem to be solved. The standardized interview carried out with a questionnaire is carried out orally, with the interviewer asking the questions and writing down the answers. It reduces the spontaneity of the subject, but increases the reliability of the results, allowing the results to be compared.
The interview versus the individual questionnaire, in which the respondent fills out the questionnaire, has the following advantages:
- The interview allows you to obtain more complete information.
- The sed interview can be applied to both literate and illiterate individuals.
- A greater number of responses are obtained from respondents than with questionnaires.
- Both the responses of the respondent and complementary information from the interviewer are collected.
- While a questionnaire received by mail can be forgotten or torn up on the spot, a person must receive it and attend to it, and not responding is more compromising.
The disadvantage is that it is expensive, since you have to have trained personnel and this is expensive.
Statistics provides the methodological tools to be able to affirm that the results obtained in an investigation are valid. However, the research loses its validity if the source information is altered, incomplete or biased. The initial stage of data collection and systematization is essential to achieve a successful result.
It is necessary to know the types of computer sources and the most used tools and resources to systematically collect the required data. There are several options and the
researcher must decide between them in order to achieve the best results. Sampling methods cannot be left out of the sight of the
expert researcher or in their infancy.