THE SURVEY AND ITS USE IN MARKET RESEARCH
Ad-hoc surveys correspond to one of the most widely used methods in market research, mainly because large amounts of data such as attitudes, interests, opinions, knowledge, behavior (past) can be collected through surveys., present and intended), as well as the classification data related to demographic and socio - economic measures. The collection of information through the surveys is carried out with the express collaboration of the individuals surveyed and using a structured questionnaire as an instrument for collecting the information.
Therefore, the survey is a procedure used in market research to obtain information through questions addressed to a representative sample of individuals from the population or universe so that the conclusions obtained can be generalized to the population as a whole following basic principles of statistical inference, since the survey is based on the inductive method, that is, from a sufficient number of data we can obtain conclusions at a general level.
The main advantage of the survey over other techniques is its versatility or ability to collect data on a wide range of information needs. However, it also has certain limitations or drawbacks such as:
- The possible resignation of the respondent to supply the information to be obtained. The respondent may be unable to provide the required information for multiple reasons (who do not remember facts, do not know them, do not distinguish between different situations, etc.). Interrogation can influence the respondents' responses due to tiredness in the interrogation when it comes to excessively long surveys, for giving socially accepted answers, etc.
These limitations or inconveniences of the survey can be avoided or reduced through an exhaustive control of the information gathering instrument, that is, by means of an adequate questionnaire design.
TYPES OF SURVEY
There are three basic methods of conducting ad - hoc surveys:
- personally, by phone or by mail.
In the personal survey, the questions are asked in a direct meeting between the respondent and the interviewer, in the telephone survey the situation is similar, except that communication is carried out by telephone and in the postal survey, respondents are asked to complete and return the questionnaire. that is mailed to them.
These different types of survey have advantages and disadvantages that determine whether their application is more recommended in certain situations. In the following table we summarize the main advantages and limitations of the personal, telephone and postal survey and in the following sections we develop in detail the most relevant characteristics of each of these survey methods.
Advantages and Limitations of Different Survey Methods
| Method | Advantage | Disadvantages |
| Personal survey |
|
|
| Telephone survey |
|
|
| Postal survey |
|
|
THE PERSONAL SURVEY
The personal survey is perhaps the most popular method and the one that has been most widely used in the collection of primary information, mainly due to the advantages that it presents compared to other types of survey. It consists of a personal interview that is established between two people, at the initiative of the interviewer, to obtain information about certain objectives.
Following the scheme proposed in Table 1, the main advantages of personally conducted surveys are as follows:
- Among the different survey methods, the personal ones provide the highest response rate, since when contacting the people to be surveyed, it is unlikely that they will decline to answer the questionnaire or will not finish it once it has started. knows with certainty who answers and the influence of third parties is avoided. Less evasive and vague answers are obtained, since the interviewer can clarify any type of doubts that may arise in the questionnaire and the typical responses of no knows does not answer They allow auxiliary materials to delve into certain topics, such as photography, prints, products, etc. Secondary data can also be obtained from the interviewee such as presence, family environment, sex, etc.
On the other hand, personal surveys present the main drawbacks:
- A high cost, since the time necessary to carry out the interview must be added the time the interviewer needs to travel to the place of residence of the person to be interviewed, which, together with the cost of travel, represents a significant cost per interview bias may be caused by influences of the interviewer, that is, the performance of the interviewer in the survey, their physical presence, their explanations, their intonation, etc. they can represent a source of distortion in the information obtained. To avoid these possible biases, it is important that in the selection the interviewers take into account certain personal qualities, honesty, objectivity, neutrality and ease for human contacts,being also important a good general training on the interview technique, forms of presentation, clarification of answers, etc., in addition to the specific instructions of the research in particular that although they deal with the same subject they use questionnaires and search procedures Different sample units. Taking into account that the research base rests on the veracity of the information collected, it is necessary to check that the data presented by the interviewers are true and that they really come from the selected sample. To verify this, an exhaustive control of the same is necessary in personal surveys, through a review of a small sample of the surveys carried out by each interviewer.In addition to the specific instructions of the research in particular, although they deal with the same subject, they use questionnaires and search procedures from different sample units. Taking into account that the basis of the research rests on the veracity of the information collected, it is necessary to control that the data presented by the interviewers are true and that they really come from the selected sample. To verify this, an exhaustive control of the same is necessary in personal surveys, through a review of a small sample of the surveys carried out by each interviewer.In addition to the specific instructions of the research in particular, although they deal with the same subject, they use questionnaires and search procedures from different sample units. Taking into account that the basis of the research rests on the veracity of the information collected, it is necessary to control that the data presented by the interviewers are true and that they really come from the selected sample. To verify this, an exhaustive control of the same is necessary in personal surveys, through a review of a small sample of the surveys carried out by each interviewer.Taking into account that the research base rests on the veracity of the information collected, it is necessary to check that the data presented by the interviewers are true and that they really come from the selected sample. To verify this, an exhaustive control of the same is necessary in personal surveys, through a review of a small sample of the surveys carried out by each interviewer.Taking into account that the research base rests on the veracity of the information collected, it is necessary to check that the data presented by the interviewers are true and that they really come from the selected sample. To verify this, an exhaustive control of the same is necessary in personal surveys, through a review of a small sample of the surveys carried out by each interviewer.
As far as personal surveys are concerned, they can be carried out either in the respondent's home or "on site". Household surveys are used in general research, guaranteeing the randomness of the selected households through different sampling methods. On-site surveys are carried out in different places related to the objective of the investigation (establishments, bus stations, etc.) in order to carry out the survey in a more realistic situation.
THE TELEPHONE SURVEY
The telephone survey is a method whose use has been increasing in recent years as the number of households with telephones increases and, therefore, their representativeness improves. Initially, it was used to conduct audience tests of programs and advertisements broadcast on radio and television, but later its use in gathering information has become widespread. From the point of view of market research, the telephone survey can be used in three ways: (1) as the only means of conducting research, (2) as a technique combined with others, that is, a survey can be designed to be carried out personally,by phone and by mail depending on the characteristics and conditions of the research and the population from which the information is intended to be obtained and (3) as an auxiliary means in the use of other techniques such as, for example, to establish appointments in the conducting personal surveys, informing or claiming the mailing of questionnaires, completing questionnaires, carried out through postal surveys, as a means of controlling the interviewers in personal or telephone interviews, etc.
Among the main advantages of the telephone survey we can highlight the following:
- It is a technique that allows information to be obtained very quickly, since in a few hours an interviewer can carry out multiple surveys. Compared to a personal survey, its cost may be lower, although this depends on various factors such as the duration of the questionnaire, the distance, telephone rates, etc. It allows easier access to busy people, especially for short surveys, which are often difficult to reach through personal surveys. In relation to the response rate, the number of people who accept answering a survey by phone is usually quite high,although in recent years the rapid expansion of "targeted" telemarketing is causing a decrease in the public's willingness to cooperate in telephone surveys as a result of the increase in telephone calls, often unwanted, and the confusion that can occur in individuals between telemarketing and market research studies. For this reason, it is essential to give a clear and convincing explanation that this is a commercial research study, with no intention of selling anything, to inform you of the objective of the research, who is conducting it and how that person has been selected.It is essential to give a clear and convincing explanation that this is a commercial research study, with no intention of selling anything, to inform you of the purpose of the research, who is conducting it and how that person has been selected.It is essential to give a clear and convincing explanation that this is a commercial research study, with no intention of selling anything, to inform you of the purpose of the research, who is conducting it and how that person has been selected.
The main drawbacks or limitations of the telephone survey are:
- The possible problems of lack of representativeness of the sample as a result of which only people who have a telephone can be surveyed, although in the event that the universe is made up of companies, this problem does not exist, since practically one hundred percent of them have a telephone. Regarding the selection of telephone numbers, it is convenient to generate them randomly through a telephone directory as there are numbers that are not registered and do not include new subscribers. Another aspect that is a limitation of this method is the brevity of the interview. which is recommended not to exceed 15 minutes. This implies that the questionnaire is also brief and, therefore, not too much information can be collected, not being able to use auxiliary or exhibition material, such as cards, photographs, sketches, etc.It also greatly limits the type of information that can be obtained through electronic surveys.
Regarding the design of the questionnaire for telephone surveys, we must bear in mind, in addition to the general rules that we will develop in section 3.5, the following aspects:
- The brevity of the interview requires that the questionnaire be shortened and does not contain too many questions. The questions should preferably be closed or semi-open and with a range of responses that are not very wide. Multiple-response questions must be eliminated or broken down into successive questions.
THE POSTAL SURVEY
The postal survey consists of sending a questionnaire by mail to the people who make up the sample in the hope that they will return it completed by the same means. As an interview technique, it can sometimes be used essential when it is the only means of establishing contact with the respondents and in other situations it can be used as an optional technique depending on the characteristics of the research. Among the main advantages of the postal survey we must highlight.
- Its low cost compared to the other two survey methods, since it is an economic system for collecting information, especially when the geographical scope of the research is national or international. It is a more flexible system for the interviewee who can answer the questionnaire in the most convenient time and using the time you want, which can lead to more thoughtful responses. Possible biases that may be caused by the influence of the interviewer in the process of conducting the survey are avoided.
Regarding the disadvantages and disadvantages of the postal survey, we highlight as most important:
- The low number of responses received in relation to shipments made, which is around 10% or 20%. In order to deal with this problem of the low response rate, it is necessary to establish a series of motivating mechanisms, such as, firstly, that together with the questionnaire, a letter of representation be sent out, highlighting the need for the collaboration of the respondent and the importance of your answer for the purposes of the investigation. Secondly, another mechanism that increases the response rate is to offer a material incentive, in addition to the moral incentive that is intended to be given with the letter. In this sense, various forms have been used, such as, for example, offering a participation in raffles to those who respond, gifts, discount vouchers on certain products,facilitate the results of the investigation, etc. The identity of the respondent is an uncontrolled variable, since we cannot guarantee that the recipient of the questionnaire actually answers the questionnaire and, at the same time, there may also be influences from other people on the responses of the respondent and This, logically, introduces distortions in the sample. Postal surveys are limited to situations in which there is a directory of the universe to be investigated, which is not always frequent in investigations. Another problem that motivates the absence of a The interviewer refers to the small amount of information that can be obtained through the questionnaire, which must necessarily be brief, unless the respondents are emotionally involved in the subject to be investigated.Although there is no general rule on the length of the questionnaire, its brevity usually increases the number of responses. Finally, we must point out the problems associated with the representativeness of the sample, since the lack of response may lead to the structure of the real sample that is obtained from the questionnaires answered does not fit the structure of the defined sample, being necessary to either send new questionnaires to the strata of the sample from which we have not received enough answers, or complete the answers necessary by means of telephone or personal interviews or to discard a certain number of surveys of those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.Normally, its brevity increases the number of responses. Finally, we must point out the problems associated with the representativeness of the sample, since the lack of response may lead to the structure of the real sample obtained from the questionnaires answered. does not conform to the defined sample structure, and it is necessary to either send new questionnaires to the strata of the sample from which we have not received sufficient responses, or to complete the necessary responses through telephone or personal interviews, or to discard a certain number of surveys of those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.Normally, its brevity increases the number of responses. Finally, we must point out the problems associated with the representativeness of the sample, since the lack of response may lead to the structure of the real sample obtained from the questionnaires answered. does not conform to the defined sample structure, and it is necessary to either send new questionnaires to the strata of the sample from which we have not received sufficient responses, or to complete the necessary responses through telephone or personal interviews, or to discard a certain number of surveys of those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.We must point out the problems associated with the representativeness of the sample, since the lack of response may lead to the fact that the structure of the real sample obtained from the questionnaires answered does not conform to the structure of the defined sample, being necessary or Either send new questionnaires to the strata of the sample for which we have not received enough responses, or complete the necessary responses through telephone or personal interviews, or discard a certain number of surveys from those strata of the sample whose responses were more numerous, In the latter case, the sampling error increases.We must point out the problems associated with the representativeness of the sample, since the lack of response may lead to the fact that the structure of the real sample obtained from the questionnaires answered does not conform to the structure of the defined sample, being necessary or Either send new questionnaires to the strata of the sample for which we have not received enough responses, or complete the necessary responses through telephone or personal interviews, or discard a certain number of surveys from those strata of the sample whose responses were more numerous, In the latter case, the sampling error increases.since the lack of response may lead to the fact that the structure of the real sample obtained from the questionnaires answered does not conform to the structure of the defined sample, being necessary to send new questionnaires to the strata of the sample of those who We have not received enough responses, or complete the necessary responses through telephone or personal interviews, or discard a certain number of surveys from those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.since the lack of response may lead to the fact that the structure of the real sample obtained from the questionnaires answered does not conform to the structure of the defined sample, being necessary to send new questionnaires to the strata of the sample of those who We have not received enough responses, or complete the necessary responses through telephone or personal interviews, or discard a certain number of surveys from those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.or complete the necessary responses through telephone or personal interviews, or discard a certain number of surveys from those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.or complete the necessary responses through telephone or personal interviews, or discard a certain number of surveys from those strata of the sample whose responses were more numerous, in the latter case increasing the sampling error.
As noted above, the sending of a postal questionnaire must be accompanied by a cover letter with which it is intended to stimulate respondents to complete and return the survey and, at the same time, serves as a means of communication between the Investigator and the respondent to make known the reasons for the research, who is doing it, its objectives, etc.
In the presentation and writing of this letter, the basic principles of direct advertising must be kept in mind, that is,
- should be normally addressed to the recipient; written from the point of view of the recipient and not the sender; the beginning and the end of the letter are the two most important parts, so attractive aspects should be included, avoiding unnecessary introductions and long goodbyes final; must be signed in original and must not be excessively long, it is normally recommended that its length does not exceed one page.
Regarding the content of the letter, express reference should be made to the following aspects:
- who conducts the research, objectives and repercussions, collaboration needs of the interviewee and anonymity of the respondent and overall treatment of the data provided. In Annex 1 we collect an example of a cover letter made for a postal survey that was carried out at the Department of Economics and Business Management of the University of Las Palmas de Gran Canaria.
The questionnaire, both in its form and in its content, must be easy to answer, the questions must be formulated in a clear and concise manner, and must be closed or semi-open, ensuring that the answer is not ambiguous. Brief, clear and precise instructions should be included as an integral part of the questionnaire or in a separate document, to facilitate its completion. It is also convenient for the questionnaires to be numbered, so that if no response is obtained after a first submission, it can be repeated by sending a second questionnaire. Obviously, the submission must also contain a stamped envelope with the address of the researcher for the return of the completed questionnaire.
CRITERIA FOR SELECTING THE TYPE OF SURVEY
Once the previous considerations on the different methods of capturing information through ad - hoc surveys have been exposed, the problem of the ideal choice for research depends, logically, on the objectives of the research and its adequacy based on the advantages. and limitations that we have indicated for each of the methods. Thus, for example, establishing the versatility, the time to obtain the information, the cost, the precision and the convenience for the respondent as criteria, we can rank the choice of these three methods as shown in Figure 2.
In relation to versatility, factors such as the greater or lesser number of questions that can be included in the questionnaire, the amount of data that can be obtained, the type of questions and the possibility of using visual aids such as cards, photographs, etc. They determine that the personal survey is much more flexible and allows obtaining a greater variety of information than telephone or postal surveys.
When time is a key factor in the investigation, the telephone survey, logically, is the fastest way to obtain the information, and although in the personal survey the number of interviewers can be increased to meet the time constraints, in practice The problems associated with training, coordinating, and controlling an excessively large group make it neither feasible nor economically profitable to increase the number of interviewers. In the postal survey, it is impossible to shorten the time that elapses from the time the questionnaires are sent until the completed ones are returned.
Regarding cost and in general terms, the postal survey is the cheapest, and the personal survey the most expensive. However, there may be exceptions depending on the size of the questionnaire, place of residence of the individuals who make up the sample, telephone rates, etc.
Regarding the precision of the method used, there is no doubt that there are a large number of factors that determine the greater or lesser precision of the questions, the greater or lesser willingness of the respondents to answer, etc. However, there are a number of aspects that affect the accuracy of the data and differ in each of the methods used, such as the degree of control over the sample, which refers to the ability to identify and obtain the data from a sample. It is representative of the strata of the population or universe and the ability to ensure the cooperation of the respondents contacted, control in supervision, that is, the possibility of minimizing questions in the survey that are not understood. In this sense,the personal survey is the one that allows greater control over the sample and overcomes the problems associated with the difficulties that the respondent may have when answering certain questions, while the postal survey, logically, is the one that results in the least errors derived from the interviewer himself.
Finally, the postal survey, followed by the telephone survey, is the most flexible from the respondent's perspective, since it allows you to fill out the questionnaire at the most convenient time and using the time you want.
Obviously, the problem of method choice does not only depend on these five criteria, and it is also very difficult for a market research approach to have one method that is best for all criteria. Therefore, the most appropriate will be the one that best suits the information needs of the investigation, taking into account time and cost limitations. On the other hand, we must emphasize that these three methods of capturing information through ad-hoc surveys are not mutually exclusive, but can be combined and thus obtain better results in the research design based on the advantages of each one of them..
THE PROCESS OF A SURVEY
Once the research objectives have been established, the information needs that are required and the type of survey, personal, telephone or postal, that is most convenient to use has been defined, the first phase of the survey process is the sample design. which, as we analyze in topic 3, involves deciding what the universe or population will be from which we are going to obtain the information, delimiting the sample size and selecting the most appropriate sampling method.
Secondly, the questionnaire must be drawn up, the design of which involves certain difficulties and a series of general criteria must be taken into account that refer to the content, type and sequence of questions. Once the questionnaire is prepared, it is also necessary to pretext it to a small sub-sample to detect possible errors and correct them.
Once the final questionnaire has been drawn up, the field work is organized and carried out, that is, the effective collection of data through the corresponding procedure, the type of survey carried out, that is, by personal interview, by mail, or by telephone.. In conducting the surveys, with the exception of the postcards, the interviewer must be familiar with the questionnaire, ask the questions exactly as they appear written in the questionnaire and in the same order, and it is also important to use some investigative techniques to ensure that the Respondent answers the questions. In this sense, good selection and training of interviewers and supervision of field work are of great importance, being necessary to introduce the pertinent control mechanisms. On the other hand,It is necessary to inspect the data obtained from the surveys in order to verify that the questionnaires are correctly answered, that the answers given are consistent, and reject those questionnaires that are not correct.
The creation of the database basically consists in transferring the codes of the questionnaire responses to magnetic media (disk or tape) so that they can be transferred to files and read and processed by the computer. The analysis of the information is carried out based on the objectives of the investigation, using the statistical treatment that is appropriate in each case through the existing computer programs, among which we highlight the SPSS, Gandia, BMDP, etc… The analysis technique Data can be univariate, bivariable and multivariable depending on whether it is, respectively, to analyze a single variable in isolation, the relationship between two variables or the relationship or interdependence between more than two variables.
Finally, the last phase of the research process is the interpretation of the results obtained, both from the tabulations and from the multivariate analysis techniques used. These results will confirm or refute the hypotheses raised, will allow certain conclusions to be reached and may highlight the need for further research. Finally, the entire investigation process, from the approach to the problem to be investigated, the methodology followed, the results obtained to the conclusions reached will be included in a written report that will involve the synthesis of the entire investigative process.
Figure 3 includes these basic elements that constitute the procedure for conducting a survey and then, in the next section, we develop the most relevant aspects related to the design of the questionnaire, leaving for the next topic the analysis of the following phases or stages of this process.
Table 2. Stages of the Conduct of a Survey
SAMPLE DESIGN
|
QUESTIONNAIRE DESIGN
|
| ORGANIZATION AND CONDUCT OF FIELD WORK |
| DATABASE CREATION AND INFORMATION ANALYSIS |
| INTERPRETATION OF RESULTS: FINAL REPORT |
THE DESIGN OF THE QUESTIONNAIRE
The questionnaire is the formalized scheme to collect the information of the respondents that contains the questions to ask and the spaces destined for the answers; that is, it is the translation of the informative objectives of the research into specific questions.
The design of the questionnaire is a key element in the process of conducting a survey, largely conditioned on how successful the design of the questions is. Hence the well-known topic that conducting the questionnaire is more an art than a technique, since there are no principles that guarantee the development of an effective and efficient questionnaire. The design of the questionnaire is more a technique learned by the researcher through his experience and really from this accumulated experience a series of rules or guidelines have emerged that can be very useful to design a questionnaire and that refer to the type or format of questions to use, their wording and their order or sequence.
TYPE OF QUESTIONS IN THE QUESTIONNAIRES
The basic element of the questionnaire, as clearly deduced from its name, are the questions. That is why the goodness of a questionnaire depends on the kind of questions used in it and its adequate formulation. From the point of view of market research, the questions in a questionnaire are the interrogative expression of the empirical variables or indicators for which it is interesting to obtain information through the survey. Therefore, the questions in the questionnaire are subdivided into responses, which are the elements of variation or categories of the variable to which the question refers. There is a wide typology of questions and various forms of classification;For this reason, and without wishing to be exhaustive, we will focus on the different types of questions according to the type of response and according to the function they can fulfill in the questionnaire.
Firstly, and depending on the type of response, we can distinguish the following types of questions:
- Open questions: They are those in which no type of answer is established, leaving this to the free will of the respondent. In other words, the respondent's answer is not previously defined and the interviewer limits himself or herself to recording the answer obtained to the letter. Closed Questions: They are the ones that the respondent chooses to choose one or more of the answers previously defined in the questionnaire.; the responses are known a priori and are fully precoded.
Both types of questions have their advantages and limitations. With open questions you can always discover new answers and opinions that had not been taken into account and it is possible to avoid that the answers obtained may be biased by the researcher's hypothesis line. However, to be able to quantitatively analyze these types of questions, it is necessary to group the answers and code them later, which involves some difficulty and takes time. On the other hand, they also have the disadvantage that the interviewee can answer in a line that is not of interest to the investigation or even that does not suit the question. However, they are especially suitable in exploratory research or when there is little knowledge of the possible answers.
Closed questions, on the other hand, are easier to answer since they require less effort on the part of the respondent and, logically, it is not necessary to group or code them afterwards. Therefore, they are usually the most used questions in questionnaires.
An intermediate alternative between closed and open questions and frequently used in questionnaires is the use of semi-open questions, that is, closed questions with an open item to reserve the possibility of incorporating other answers different from those previously selected.
In closed questions, in turn, we can distinguish between:
- Dichotomous questions: They are the ones with only two answers. Fan answer questions: They are those in which the respondent must choose between a certain number of possible answers. In this type of questions, when the possible answer options are numerous, it is convenient to ask them with a card instead of reading the answers in order to prevent that some answers are more likely to be chosen than others, since when the number of answers is relatively Large, the latter are more likely to be remembered than the former. This type of questions with cards is also convenient to use in questions with difficult answers or that may produce some rejection when answering them. Subjective scale questions:They are those questions in which the answers are graduated in increasing or decreasing intensity on the desired information point. That is, the respondent positions himself subjectively with respect to the different response categories. (What do you think…? Good Bad Worse) Questions of numerical subjective scale: They are similar to the previous ones but with numerical positions. Table questions: They are mainly used to obtain more than one piece of information that is normally collected in double-entry tables.They are mainly used to obtain more than one information that is normally collected in double-entry tables.They are mainly used to obtain more than one information that is normally collected in double-entry tables.
There are other types of questions that have special functions within the questionnaire, which constitute special mechanisms of inquiry or serve different purposes of information:
- Filter questions: These are closed questions, with few options (they are usually dichotomous questions) whose answers depend on whether or not to ask subsequent questions. In short, this type of question constitutes a bifurcation in the questionnaire. Control questions: They are often used frequently in questionnaires in order to check the veracity and consistency of the answers given above. They include false or cheating answers in order for the respondent to become aware of them. Consistency questions: These are similar questions to control questions that are intended to check the consistency of the interviewee's answers. These are similar questions, but worded in different ways, which are placed at a distance from each other to see if the answers to both questions are consistent.Introductory or contact questions: These are the questions asked to start the questionnaire or to move from one topic to another in order to create a climate of trust and interest in the interviewee.
DRAFTING THE QUESTIONS OF THE QUESTIONNAIRES
Regarding the wording of the questions, in general a well-formulated question is one that:
- Does not influence the direction of the response Does not incite an inaccurate response that does not correspond to the information sought.
In this sense, there are a series of rules or recommendations based on common sense and simple experience that must be borne in mind when formulating the questions and which we detail below:
- An accessible, understandable language must be used, that is, the vocabulary must be simple, direct, and familiar. It should not be forgotten that except for specific studies, the questions are addressed to people of the most diverse classes and training and consequently it must be taken into account, excluding the use of difficult words or concepts. Ambiguous words or phrases that may have different interpretations. For example, the term "frequently" in the use of a product can mean every day, two or three times a week, etc. Questions should be neutral or impartial, avoiding that in some way a certain answer may be implicit. For example,to the question, are you in favor of the Las Palmas City Council creating more green areas in the city? all the interviewees will answer positively and these answers would not coincide if the question were asked in the following way: are you in favor of creating more green areas in the city, even if this implies an increase in taxes by the City Council of Las Palmas?. Avoid using words loaded with certain connotations, since in many cases a question formulated in a certain way can suggest the answer in a certain direction. Questions that are difficult to answer or require calculations should not be included, or memory efforts on the part of the respondent and they probably won't remember them all,it is preferable to list the products or services and ask if they use them or not. They should not ask double-effect questions, that is, questions whose wording may involve a double answer.
There are questions that refer to sensitive or sensitive issues that can put the interviewee in an embarrassing situation or questions that have an element of prestige or adherence to socially accepted norms, which generally lead to answers with a high index of bias. To avoid the possible risks that may arise in this type of question or the respondent's refusal to answer them, different procedures have been developed to obtain more exact answers:
- Through counteracted statements that consist of starting the question with a statement that suggests that the behavior in question is common among the population, and then asking the question through indirect statements, for example by asking the sensitive question but referring to other people, since it supposes that the behavior or attitude of the respondent will be reflected in the answer. By asking the question with a card so that the respondent responds with the number that refers to the answer on the card. Through batteries of questions that allow us to approach the subject in an indirect way.The questions should not be excessively long since the questionnaires should be fluid and with short questions where a rhythm of questions and answers can be imposed that does not bore the respondent or the interviewer. When this rhythm is altered with a question whose text is excessively long or repetitive, the conversation loses dynamism and negatively affects the development of the interview. The wording of the questions should invite collaboration, should be formulated in a friendly, natural way, and with normal language. The interviewee should not have the feeling of being subjected to an examination, so the questionnaire must be designed in such a way that a situation similar to normal conversation occurs between the respondent and the interviewer.When this rhythm is altered with a question whose text is excessively long or repetitive, the conversation loses dynamism and negatively affects the development of the interview. The wording of the questions should invite collaboration, should be formulated in a friendly, natural way, and with normal language. The interviewee should not have the feeling of being subjected to an examination, so the questionnaire must be designed in such a way that a situation similar to normal conversation occurs between the respondent and the interviewer.When this rhythm is altered with a question whose text is excessively long or repetitive, the conversation loses dynamism and negatively affects the development of the interview. The wording of the questions should invite collaboration, should be formulated in a friendly, natural way, and with normal language. The interviewee should not have the feeling of being subjected to an examination, so the questionnaire must be designed in such a way that a situation similar to normal conversation occurs between the respondent and the interviewer.therefore, the questionnaire must be designed so that a situation that is most similar to a normal conversation occurs between the respondent and the interviewer.therefore, the questionnaire must be designed so that a situation that is most similar to a normal conversation occurs between the respondent and the interviewer.
STRUCTURE OF THE QUESTIONNAIRE
When writing the questions, special care must also be taken in the order in which they are included in the questionnaire, since the sequence of the questions may influence the nature of the respondent's responses and, therefore, skew them. For this there are a series of general guidelines that can be very useful in structuring the questionnaire:
At the beginning of the questionnaire, a presentation should be included requesting the cooperation of the respondent and specifying the objectives of the research, who carried it out, and an explicit statement that the information provided will have global treatment, since the guarantee of anonymity of a questionnaire it is essential to create a willingness to answer.
The first question must be general, simple and enhancing the respondent's interest. Sometimes this introductory question is not related to the information needs of the research, since its only objective is to achieve the cooperation of the respondent and establish a relationship of harmony with him.
General questions should precede more specific questions.
Generally, the simplest questions should go to the beginning of the questionnaire, reserving the most committed ones for the end and leaving the intermediate space for the most substantive and important ones of the investigation. In other words, the least problematic questions should be asked initially, and gradually the more complex and personal questions will be introduced.
The flow in the process of the questions should be logical according to the perspective of the respondent and, at the same time, the questions should be grouped according to their theme to avoid embarrassment of the interviewee by asking questions related to the same topic in different phases of the interview.
The identification data of the respondent, used as analysis variables and globally to check the reliability of the sample, must be formulated at the end of the questionnaire.
PERFORMING THE PRETEST
Once the questionnaire has been designed applying the previously stated rules, it must be pretested or tested among a small subsample in order to detect possible problems that may arise in the questionnaire. In this preliminary test it is advisable to use experienced interviewers to detect possible questions that cause confusion in the respondent, words that are not understood, problems that can occur in the standard categories of closed questions, etc. Also, the open response format it can be used in this pretest to determine the most appropriate or frequent answer categories and, later, turn it into a closed question.
Some of the variables that determine a poor design of the questionnaire are the number of refusals to answer questions, a high proportion of "do not know / do not answer" and also blank questions. Unless the sample has been erroneously selected and, for example, it is intended to investigate a highly qualified topic among people with little training, it is not normal that the percentage of blank questions or that they do not know how to answer exceeds 10% if the questionnaire is well designed. When these cases occur, the causes of the problems must be investigated, the errors detected must be corrected until the final questionnaire is drafted.
Regarding the number of people who make up the subsample of this preliminary test, it depends on the characteristics of the sample, although it usually tends to range between 25 and 50 people and, logically, the people who make up the subsample must coincide in their characteristics with the sample of the definitive investigation.
QUESTIONNAIRE CODING
The coding of the questionnaires is intended to facilitate the introduction into the computer of the information that is collected in them for their subsequent computer-statistical treatment. It consists of assigning to each of the questions that make up the questionnaire a different number and in turn each category of response corresponding to the different questions will also have a different number associated to identify the answer with its corresponding numbering and the interviewer is limited in In the case of closed questions, circle the corresponding code. For open-ended questions, the coding of the answers must necessarily be carried out after carrying out the field work.
In the case of multiple questions (the interviewer can answer more than one answer category, as many as indicated in the questionnaire), it will be necessary to enter as many numbers as possible answers the respondent can quote. In the event that the respondent does not cite all the possible responses contemplated in the questionnaire, the rest of the response codes will be completed with zeros. Another feasible alternative in coding multiple choice questions is to consider each category as a dichotomous YES / NO question (1/0). In Annex 2 we collect an example of a questionnaire where its coding can be perfectly appreciated.
Once the field work has been carried out, which we will analyze in detail in the next section, it is necessary to create the database, which consists of entering the answer codes of the questionnaires in a computer format through a computer program or package. (Gandia Bardwin). Once all the data has been entered and recorded, and as a previous phase to the analysis of the information through the pertinent statistical treatment, it is convenient to carry out a simple tabulation of all the variables, that is, count the number of cases included in each one of the categories contemplated in the answers of each question through a simple frequency of each variable, in order to verify that the database does not contain errors.
FIELD WORK
Field work is that stage of the research project during which the interviewers contact the respondents, administer the questionnaires to collect the information, record the data and return the questionnaires for later control and processing.
The planning of the field work is a basic aspect in the process of a market research and its development depends on the method of obtaining the information that is used; thus, fieldwork for a study based on personal surveys is totally different from that of research based on postal surveys.
RECOMMENDATIONS IN THE INTERVIEW PROCESS
In personal and telephone interviews, the interviewer can be a source of errors related to (a) lack of harmony between the interviewer and the respondent, (b) the incorrect formulation of the questions, (c) errors in recording the answers and (d) falsification of the surveys. Below we briefly analyze the most relevant aspects that refer to the four elements previously outlined to obtain better results in the interview process.
Harmony between the interviewer and the respondent
During an interview, the respondent's perception of the interviewer can directly affect the latter's ability to establish adequate harmony, since interviewers who can establish effective relationships with interviewers are able to collect more complete and accurate information.
In the case of the personal interview, the interviewer's dress and composure are important to establish a good harmony, since most people consider them as indicators of people's attitudes. In this sense, the conduct of the interviewer must be very kind and the respondent must see the interviewer as a person capable of understanding their points of view.
Formulation of questions
A series of appropriate guidelines have been developed for most of the interview situations that we briefly review below.
- Be very familiar with the questionnaire. The interviewer must study the questionnaire question by question to avoid making mistakes in their reading. The questions should be read in the most natural way possible and in a conversational tone. Formulate the questions exactly as they appear written in the questionnaire. The interviewer should not make changes to the formulation of the questions, such as stopping reading part of the question, changing or adding words, etc., as it may distort the result of the investigation. Formulate the questions in the order they appear. in the questionnaire. In the design of the questionnaire, the sequence of the questions was established so that there is a sense of continuity in the topic discussed and biases in obtaining information are avoided.Formulate each of the questions specified in the questionnaire. The respondent's response to one of the questions can answer another of the questions that appear later in the questionnaire. In this situation, the interviewer should not omit the question that seems to have been answered previously, since it may be a control question. Use investigative techniques to get the respondent to answer the question. The questions have been designed for all the respondents who are part of the sample to answer; however, sometimes it may happen that the respondent does not understand or misinterprets the question, refuses to answer it, or gives incomplete or ambiguous information. In these cases, the interviewer must use certain neutral investigative techniques such as the following:The respondent's response to one of the questions can answer another of the questions that appear later in the questionnaire. In this situation, the interviewer should not omit the question that seems to have been answered previously, since it may be a control question. Use investigative techniques to get the respondent to answer the question. The questions have been designed for all the respondents who are part of the sample to answer; however, sometimes it may happen that the respondent does not understand or misinterprets the question, refuses to answer it, or gives incomplete or ambiguous information. In these cases, the interviewer must use certain neutral investigative techniques such as the following:The respondent's response to one of the questions can answer another of the questions that appear later in the questionnaire. In this situation, the interviewer should not omit the question that seems to have been answered previously, since it may be a control question. Use investigative techniques to get the respondent to answer the question. The questions have been designed for all the respondents who are part of the sample to answer; however, sometimes it may happen that the respondent does not understand or misinterprets the question, refuses to answer it, or gives incomplete or ambiguous information. In these cases, the interviewer must use certain neutral investigative techniques such as the following:
- Repeat the question. A very effective approach is to repeat the question exactly as it appears on the questionnaire. Expectation break. If the interviewer remains silent or pauses, it creates an effective key to make the respondent aware that a more complete response is expected. Repeat the respondent's response. Respondents are often encouraged to make additional comments after repeating what they have said. This can be done while the interviewer records the answer in the questionnaire. Reassuring the respondent. If the respondent hesitates to answer, it can be very useful to make neutral comments such as the following: «there are no correct or incorrect answers,we just want their ideas on the topic ”or“ we are just trying to get to know people's ideas on this topic. ”Neutral questions or comments. When the respondent's response is somewhat incomplete, neutral questions can be used, such as "something else?" "Any other reason?" Could you tell me more than you think about this? "What does it mean?" Request more detailed clarification. By asking the question "I'm not quite sure what you mean by that, could you explain it a little better?" The respondent may feel like cooperating with the interviewer and giving more detailed information about their response..The interviewer must record the responses provided by the respondent in an unbiased manner so that it can be accurately interpreted by the coders. For this, it is necessary to record the answer at the same time that the respondent expresses it, use the same words of the respondent, do not summarize or paraphrase the respondent's responses, include comments in parentheses, maintain the respondent's interest by repeating the response when writing, use legible handwriting and in case of blank questions, explain the reasons for not answering the question. In order to avoid that survey falsifications may occur by the interviewers, it is necessary to introduce some control mechanisms for the interviewers that allow us to guarantee the goodness of the field work carried out.In the next section we will analyze the most commonly used control systems to ensure that the information obtained has not been falsified
FIELD WORK CONTROL
Bearing in mind that the basis of all market research rests on the veracity of the information collected, it is necessary to check that the data is true and that it actually comes from the selected sample. To verify this, in personal surveys it is necessary to carry out an exhaustive control through a review of a small sample of the surveys carried out by each interviewer. The commonly used control systems can be direct or indirect.
Direct control systems are based on personal or telephone control, conducting the survey again or asking if the survey was actually carried out on the indicated date. The percentage of surveys that it is recommended to control in person or by telephone ranges from 10% to 15%, depending on the confidence that the pollsters have.
Indirect control systems consist of an internal and comparative analysis of the questionnaires completed by each interviewer. Among the most used methods we can highlight the following:
- Control of stability curves. It is established through the tabulation of an interviewer's questionnaires in sequential order of carrying out their work. A high number of responses in the same sense for some questions, higher than those given by other interviewers, means a stricter control of their work. The consistency control of questions. It is checked if there are the same answers or absence of the same in the questionnaires of the interviewers. It is verified through the questions designed for this purpose to control the degree of acceptance that should be given to the information obtained.
REPORT ON THE FINDINGS OF THE INVESTIGATION
The last stage of an investigation is the final writing of the report, that is, a written document that includes the different stages covered during the investigation process and the results and conclusions drawn from the analysis are presented, providing the data, analysis and results of organized and consistent way.
Logically, there is no single report model, since it depends on the type and depth of the research carried out, the methodology used, the objectives set and the recipient of the report. From this last point of view, two types of reports can be distinguished: (a) technical reports, which are intended for people with high technical training, for which the details regarding the methodology used are particularly relevant: (b) informative reports They are intended for a more general audience, often unfamiliar with technical details, so that in this type of report the technical details are much less interesting than the conclusions and recommendations it contains, since it is assumed that they will be based on this information the decisions. Both approaches should be combined in reports,giving greater importance to the informative aspect when the results of the report are intended for company executives.
Although there is no unique and specific format that defines the structure of a report, in general terms we can establish a common structure as the most frequently used:
- Cover page. The cover should contain a title that summarizes the essence of the study, the date, the name of the person presenting the report and the name of the organization to which the report is addressed. Contents Index. In order to help readers find specific sections that may be of most interest, the index sequentially lists the topics found in the report, along with page references. On some occasions, a table index (figures, illustrations, charts) is also used to list the titles and page numbers of all visual aids. Introduction. The purpose of the introductions is to provide the reader with the basic information (background) necessary to understand the rest of the report.The nature of the introduction is conditioned by the diversity of the audience and their familiarity with the research project. The more diverse the audience, the more extensive the introduction will be. In general terms, the introduction should clearly explain the nature of the problem and the objectives of the research. Methodology. The sections that make up this section detail the sources of information used, the aspects related to sample design and the method used to obtain information. Ultimately, the methodology section should (a) summarize the technical aspects of the research project in an understandable style and (2) allow to develop confidence in the quality of the procedures used. Results.The body of the report is made up of the results of the investigation, which must be structured according to the objectives pursued. In other words, the results should be structured in a logical flow of information and we should not limit ourselves to offering a series of response frequencies, one for each question in the questionnaire. This would reflect a lack of creativity and commitment. Tables, graphs and illustrations are often used in the presentation of results to explain the different relationships between the various data analyzed. Conclusions and Recommendations. The conclusions try to summarize the main results that deserve special attention and, based on these conclusions, make the most pertinent recommendations. Annexes or Appendices.Supporting materials that are not absolutely essential in the body of the report are included: Copy of questionnaire or information collection format, calculations of the sampling plan, statistical calculations, which should not be included in the body of the report, statistical tables that are not are relevant, etc. Bibliography. You must include all the publications or sources consulted in the investigation
