| dc.description.abstract | Quite frequently a researcher is unable to evaluate the degree of truthfulness of the data obtained, while rare checking analyses show that some informations gathered in various researches are burdened with serious errors.
A detailed description of the process of data gathering by means of scheduled interviews is presented, the basis of the description being the observation of investigations conducted and the analysis of research designs and handbooks. In a communication process concerning one topic, the following messages have been distinguished: (a) a question in a schedule formulated by researcher together with accompanying instructions, (b) a question as put by an interviewer to a respondent, (c) respondent’s answer, (d) registration of that answer by interviewer, (e) classification of the records by the coder according to researcher’s instructions.
A series of statements are formulated concerning the correct course of data gathering activities in this process. So, the analysis of schedule questions has led to the conclusion that such a question is not always a request of information in the sense accepted in the logical theory of questions, but a request of an utterance of a certain type which is defined more or less precisely. A researcher expects also, from his respondents, such utterances as cannot be considered to be answers proper for the contents of the corresponding question. Hence, it is appropriate to distinguish: (a) answers desired by the researcher which are not always the proper ones in the logical sense, (b) undesired answers which should of course be as rare as possible. Both desired and proper utterances are not always sentences in the logical sense, i.e. they cannot be always considered true or false.
Logicians distinguish open and closed questions. Among the former, several types can be distinguished: (a) imperfect closed questions, (b) questions demanding the choice of an alternative as well as a set to which that alternative belongs, (c) questions demanding a narration, i.e. the choice from many alternatives belonging to many sets, (d) questions demanding the choice from an indefinite number of alternatives belonging to an indefinite number of sets.
A question, being the demand for an utterance of a definite kind, suggests more or less strongly a more or less definite kind of an answer. Actual utterances of respondents usually deviate, in respect of their contents, and, especially, of their form, from these suggestions not infrequently in accordance with researcher’s expectations. The record of such answers often extracts from them some aspects or elements only. This is so, first of all, in the case of a record categorizing the answers and, to smaller extent, in the case of the registration of the contents of answers; it does not apply to the cases when answers are recorded verbatim.
The division of questions into closed and open ones does not coincide with their division according to the categorizing or registering records used. In addition to closed questions with a categorizing record, and open ones with a registering record, it is possible to find rare cases of closed questions, with a registering record, and very frequent ones of open questions with a categorizing record. This tetramerous division seems to be more appropriate then a rather ambiguous division into closed and open questions, found in many handbooks of research methods.
Every record, together with the corresponding introductory sentence, constitutes an answer to interviewer’s own question concerning that record. In the case of the verbatim registering record, the question is: what was the verbatim answer of the respondent R to the question Q? In the case of the record registering the sense of respondent’s answer, the question is: what was the sense of the answer of the respondent R to the question Q? In the case of the categorizing record: to which one or ones from among the following categories (enumeration follows...) does the answer of the respondent R to the question Q belong? The answer to the interviewer’s own’ question, concerning the record, must be true; otherwise, the communication process is disturbed at this stage.
Classification and symbolization of the recorded answers, performed by the coder according to the researcher’s instructions, serve as a basis of computations and therefore they must be comprehensive and disjunctive. A symbol, noted down on a questionnaire or on a special card, constitutes an abbreviated record of the answer to the coder’s question: to which, from among the following classes (enumeration follows...) does the recorded answer of respondent R belong? The answer to this question is a sentence, true or false. If it is false, the process of communication is disturbed.
The records of answers to one question are sometimes classified several times, from different view-points. As a results of such classifications, noted down are several symbols relating to one question, answer and its record. The process of reciprocal communication in such a case splits into several processes. It is also possible to join several classifications and to put down one symbol instead of several ones already noted.
Then, in the course of computations, the distinctness of all the underlying communication processes is completely obliterated.
The results of researches embrace, however, not only the numerical outcomes but also their interpretation. In such interpretations the researcher gives to these outcomes a new meaning: he refers them not to the classification symbols representing the answers of the respondents, but to the phenomena studied. Thus, his interpretation relates to research units distinguished from among various phenomena studied, and is necessary for the solution of the problem posed.
In this way, the interpretation of the numerical outcomes determines the research units and their corresponding units of information. These units of information constitute a necessary logical premise in the process of transition from the records of particular answers with the corresponding symbols, to the solution of the problem. The units of information are also answers to the questions which may be called basic unitary questions of the researcher. They are always closed questions with a unique alternative. Answers to such questions correspond to classes in the final classification.
The answer to a researcher’s unitary question is always either true or false. When it is true, the researcher has been successful in the corresponding fragment of his research; when it is false, the results of his research, in that fragment, is erroneous. If the answer has not been obtained, the research in that fragment in a failure.
Researcher’s unitary questions constitute the last, logically necessary, link in the process of communication concerning one topic, and based on interviewing. It is in respect of the value of this link that some of the preceding links may be evaluated and especially respondents’ answers to the questions put to them. An undesired answer leads always to a failure in the corresponding fragment of the research. When a respondent’s answer is a desired one it is possible for the researcher to ask the basic, unitary question. But if, and only if, the answer to the researcher’s question is true, the respondent’s answer has been valid. A respondent’s valid answer must have certain attributes. Most often it is a true sentence about the phenomenon studied. However, it may be also an indicator of the phenomenon as well as a true sentence about an indicator of this phenomenon. Sometimes such an answer happens to be the phenomenon studied itself.
When a respondent’s valid answer is a true sentence about the phenomenon studied, the assumption about the truthfulness of that answer constitutes a necessary premise in the reasoning leading to recognition, as true, of one from among several answers to the basic unitary question of the researcher. When the valid answer of respondent is an indicator, such a premise is a statement that this indicator is valid. When the researcher treats respondents’ answers as the phenomena studied, his approach is characterized by cognitive minimalism consisting in the resignation from the study of facts transcending the reality evoked by the research itself.
Various factors causing deformations of the communication process are distinguished: those connected with question, an. interviewer, a respondent; with an immediate surrounding of interview and its general situation; with a place of question in an interview as well as a place of interview in a series of interviews. The existence of deformations, being a result of the influence of those factors, can be ascertained only when the information sought is known. As the source of deformations resides most frequently in the respondent giving unvalid answers, his psychical processes evoked by interviewing are analysed: distinguished are learned and unlearned reactions, various phases in the process of giving the answers as well as various levels at which this process is running. A typology of psychic processes of a respondent, from the view-point of validity of his answers, distinguishes necessary, neutral and harmful processes. | |
