Structured and Unstructured Observations Research30th January 2021
Observation may take place in the natural or real life setting or in a laboratory. Observational procedures tend to vary from complete flexibility to the use of pre-coded detailed formal instrument. The observer may himself participate actively in the group he is observing or he may be an observer from outside or his presence may be unknown to the people he is observing.
Structured observation consists in a careful definition of categories under which the information is to be recorded, standardization of conditions of observation, and is used mostly in studies designed to provide systematic description or to test causal hypothesis.
The use of structured observational technique presupposes that the investigator knows what aspects of the situation under study are relevant to his research purposes and is in a position therefore to develop a specific plan for making and recording observations before he actually begins the collection of data. Structured observation may be employed in the natural field-setting or a laboratory-setting.
Structured observation, in so far as it is used mainly in studies starting with relatively specific formulation, normally allows for much less freedom of choice with respect to the content of observation than is allowed in unstructured observation. Since the situation and the problem are already explicit, the observer is in a position to set up in advance the categories in terms of which he will analyse the situation.
The categories are clearly defined to provide reliable data on the questions to be asked. Of course, such a definition of categories is the end-product of the researcher’s efforts at trying to solve specific coding problems.
To start with, the researcher may be faced with a large number of categories. It is important that the researcher decides upon an appropriate frame of reference for categorization and trains observers accordingly.
- E. Bales has developed a procedural system of categories for recording group interaction. He has proposed 12 standard behavioural categories applicable to a wide range of group situations. Behaviour of any group member is coded in terms of careful definition of each category.
The problem of recording observations during a structured observation. The most commonly used system of recording is one that provides the observer with a number of duplicate sheets containing the list of categories to be coded.
Mechanical recording instruments have been used in some studies. For example, Chapple devised an international chronograph. Helen has developed an audio-introspect meter. Bales and Gerbrands have devised an interactional recorder. All these devices are meant to facilitate recording of observational data according to a specific principle of categorization.
Sound recordings and motion pictures have been used when it is necessary to describe the overall nature of an event or to code certain action of a member in terms of a frame of reference provided by the entire event. Of course, each of these has obvious limitations.
Although such devices as motion pictures, tape-recording and television may be very helpful in affording an overall view of a social event, their use does not by itself solve the problem of gathering data for systematic purposes.
Relevant categories for recording behaviour must be established, time-units decided upon, methods set up for recording as to who initiated an action and who was the target. In sum, if the data are to be useful for research, they must be recorded in terms of such a formal scheme.
This problem is effectively tackled by ensuring some kind of a standardization in the observational instrument. There are, however, some special problems in achieving reliable and valid observations.
These are as follows:
(1) One problem derives from the inadequate definition of the kinds of behaviour that are to be accepted as corresponding to a given concept. For example, if the concept of adjustment was not operationally defined, different observers may be inclined to regard different kinds of behaviour as empirical referents of the concept.
(2) Another factor that may lower the reliability of even a well-trained and skilled observer is the degree of confidence one must have in one’s judgement before marking a given category. For example, observers may assign the same observational items to different categories because they may themselves manifest different tendencies to perceive evidence of a particular behaviour.
(3) The constant error introduced by the observer because of the distortion of his perceptions (for various reasons) is one of the major sources of unreliability.
(4) The load of work can also hamper reliability. The result of overloading is often that the observer cannot record all relevant data and may unwittingly record some aspects rather inadequately, thus, introducing bias.
As was suggested earlier, reliability can be increased by careful training of observers. A well-developed observational procedure can be damaged by differences among different observers or by failure to understand the rules for its use. It is necessary, therefore, that a good period of time be devoted to train the observers.
Such a training entails several phases:
(i) Explanation of purposes and theory in the given study,
(ii) Explanation of categories and the rules for their use,
(iii) Purpose of each category for a theoretic scheme, and
(iv) Practice by observer-trainees, discussion on concrete difficulties and reliability-test of observers.
It should be remembered that all this may not always eliminate the constant bias shared by two or more observers. In such a case, the bias can be minimized by same events.
Lastly, we need to consider the relation of the observer with the observed. The observer must carefully prepare his entry into situation and make sure that all members of the group are willing to accept him. Since usually the observer is conspicuously engaged in recording behaviour, using timing device and other technical aids, it is barely possible to disguise the fact that he is doing research.
Hence, it is all the more important that he obtains the group’s full agreement to the inquiry.
The entry of an observer into the group, however unobtrusive, may introduce a new variable into the situation and this may change the behaviour being observed. For example, in a children’s group, the presence of adult observer may have a great distorting influence.
It is important that some thought is given to ways in which the observer’s presence may influence the outcome of research and to develop the techniques that would reduce this possibility. On the whole, people seem to get used to observers if the behaviour of the observer convinces the subjects that he means no ill.
The participant and the non-participant types of observation. This conceptual typology was introduced to social sciences by Prof. Edward Lindeman. Lindeman was very critical of studies based upon schedules of questions for which the investigator found answers by making inquiries of persons.
Lindeman considered as absurd any attempt to avoid bias by posing questions requiring a simple ‘yes’ or ‘no’ reply in a study dealing not only with the ‘what’ of life but also with the ‘why’ and ‘how’ of life. Lindeman was of the opinion that if one wished to know what the subject was really doing one should watch him and not ask him.
Nels Anderson was a intimate participant in the life of ‘Hobos’, on the road, in lodging houses and in their various activities. The tremendous insight which Anderson developed through such an exercise is amply evidenced in his study entitled ‘The Hobo.’
Participant observation has a reference to the observer sharing to a greater or lesser degree the life of the group he is observing. This sharing may be intermittent but active contacts at close proximity do afford an intimate study of persons.
W.F. Whyte in the course of his study published as ‘The Street Corner Society’ was intimately associated with the various aspects of the activities of members in Cornerville. Paul Cressey in his study entitled ‘Taxi Dance Hall’ employed the technique of participant observation and his investigators became part of the social world of the Taxi Dance Hall to the extent it was possible.
The non-participant observation, in contradistinction, is characterized by a relative lack of participation by the observer in the life of the group that he is observing. In sum, to quote John Madge, “When the heart of the observer is made to beat as the heart of any other member of the group under observation, rather than as that of detached emissary from some distant laboratory, then he has earned the title of participant observer.”
In other words, the participant observation is an attempt to put both observer and observed on the same side by making the observer a member of the group so that he can experience what they experience and work within their frame of reference.
On the contrary, the non-participant observation involves the espousal by the observer of a detached role of the observer and recorder without any attempt on his part to experience through participation that which the observed experience.
The unstructured observation is diametrically opposed to the structured observation in its ideal-typical formulation. The structured observation is characterized by a careful definition of the units to be observed, information to be recorded, the selection of pertinent data for observation and standardization of conditions of observation.
The unstructured observation represents ideally a contrasting situation in respect of all these.
(a) What should be observed? In highly-structured studies, the well-formulated research-problem or hypotheses clearly point to what data will be most relevant.
But in exploratory studies the observer does not know in advance which aspects of the situation will prove relevant. Since unstructured observation is mostly used as an exploratory technique the observer’s understanding of the situation is likely to change as he goes along.
This, in turn, may call for changes in what he observes. It should be noted that such changes called for in the foci of observation are often desirable. Such shifts in focus according to the exigencies of the situation is a characteristic of unstructured observation.
That is, the unstructured observation is flexible, it allows for changes in focus from time to time if and when reasonable clues or doubts warrant such changes with a view to facilitate taking stock of the new observational items that appear to be pertinent or important at different points in time. The observer is always prepared to draw his clues from unanticipated events in an attitude of alert receptivity.
While no stringent criteria or hard and fast rules can be laid down as to how the observer will go about observing a particular situation it would be helpful, however, to indicate some of the significant aspects that the observer can overlook only at his peril.
(1) The observer should see who the participants are, how many they are and how they are related to one another.
(2) The observer should understand the ‘setting.’ He should know in addition to its overt appearance, the kinds of behaviour it encourages, discourages or prevents and its social characteristics.
(3) The observer should also understand the purpose which has brought the subject-participants together, the nature of the purpose and how the goals of participants are related.
(4) The observer must also understand what the participants do, how, with whom and with what they do it. For example, the observer should know what stimulus initiated the behaviour, what the goal is towards which the behaviour is directed, what are the qualities of the behaviour (duration, intensity, etc.) and what are it consequences?
It should be noted that in a practical situation, it is often not possible to obtain enough clues to allow such a comprehensive description. It may also be that the course of events is too fluid to permit consideration of all dimensions of a social situation or that a certain aspect of an occurrence may be so important as to need the entire attention of the observer.
(b) Recording an observation involves two major considerations, viz:
(i) When should the notes be taken, and
(ii) How the notes should be kept.
The best time for recording is on the spot and during the event. This results in minimizing selective bias and distortions of memory. There are, however, many situations in which note taking on the spot is not feasible because this is likely to affect the naturalness of the situation and create suspicions in minds of the persons being observed.
Constant note taking may also affect the quality of observation, as the observer has to divide his attention between observing and writing. In consequence, during the process, the relevant aspects of the situation may be lost to the eye.
In a situation where on the spot detailed note taking is not possible, the memory of the observer may be too heavily taxed if recording is postponed to the expiry of an observational period. In certain situations, it may also help if the observer retires from an on-going situation for a few minutes every hour to make more detailed notes. It is important that the observer should pen down as soon as possible, after the period of observation, a complete account of everything important in the situation. The facility of recording improves if the observer evolved some kind of indexing system.
(c) Ensuring the accuracy of observation is another important concern of the observer. In situations where for some reasons, immediate recording is not possible, he is likely to find that by the time he sits down to write his observations; his memory does not accurately feed in the relevant details.
In order to check the accuracy and completeness of the record, the observer should, if feasible, compare it with a record made by a tape recording equipment. Of course, this is not always feasible; besides, tape recording captures only the auditory stimuli in the situation.
The next best solution is to have two or more people observe the same event. They can later compare their notes and check bias. This is an excellent way to discover one’s blind spots. Two observations may be qualitatively different; against this, two observers from different backgrounds may be employed to observe the same situation. This is understandably a limited remedy.
It happens quite often that the observer injects an overdose of interpretation into his records. This may adversely affect the validity and reliability of his conclusions. One way out of this is to have two observers record the same vent using the same system. A subsequent comparison, between their records may go some way in detecting the intrusion of interpretation.
The participant observer, by virtue of his typical position, faces formidable difficulties in maintaining baselessness. Such an observer may get involved emotionally with some of the people he is studying. This affects his objectivity.
To gain access to intimate data, the observer may allow himself to be absorbed into particular situation he is studying. But this very factor may make him to accept uncritically the behaviour that he should be trying to explain. This problem can be met mainly by the observer becoming aware of his proneness or tendency to take things for granted. An outsider serving as a check may bring home to the observer his blind spot.
It is also possible to detect blind spots by breaking up or dissecting the perceptual field so the factors that lead it to be seen in a particular way lose much of their force. In other words, by approaching the situation in an analytical way the observer may be able to lessen the distorting influence of certain factors that are likely to lead to bias.
The natural way of seeing the situation is to see the action as one centred around the principal actors. But an inconspicuous person, seemingly very insignificant in the situation, or sometimes even a dead person, may be the real center of the situation (e.g., in ceremonies dealing with the propitiation of the soul of a dead person).
An effective screw to control accuracy in observation and interpretations is for the investigator to establish a sort of relationship with the subjects which makes it possible for him to take them into his confidence about the research.
A participant observer’s situation is likely to create inner conflicts within the investigator. This, in turn, may interfere with objectivity. Should the group being observed be undergoing an emergency of some kind, there is indeed a strong pressure on the observer to become an active participant.
He may have to abandon at least temporarily, his detached position as an observer. But if he does enter into the center of activities of the group, he risks the danger of losing his identity as a scientist. Thus, the participant observer is in a dilemma; resulting either way, in the loss of objectivity.
Rosenfeld suggests that bias arising from inner conflicts may be minimized if one is aware of the conflicts and of the nature of one’s defence.
The final issue relates to the relationship between the observer and the observed. In field observation faulty approach vis-a-vis the subjects may have dire consequences for the inquiry. Since the method is applied in the actual life-sphere of the persons, the observer’s mistakes cannot remain insulated incidents.
The observer must decide before he approaches the potential subjects, whether to reveal the facts that he is a researcher or to enter the situation under some other guise. There are advantages as also disadvantages in both these approaches.
It may for certain reasons seem preferable to make known to the subjects his real role as the researcher. This approach is relatively simple compared to disguised observation. Secondly, it increases substantially one’s opportunity to get information which he would get only very indirectly were he to approach them in disguise.
Thirdly, the open declaration approach does not hold the possibility that his activity will harm any of the people in the situation whereas the disguised observer must consider this possibility seriously.
The obvious disadvantage of a direct approach is that this may make the subjects conscious only to the detriment of naturalness of behaviour the observer wants to observe. The researcher therefore has to weigh carefully the relative gains and losses of these two approaches before employing any one.
Entrance into a community requires a very careful staging. If there are many more than two sides to be approached simultaneously, the issue becomes all the trickier. The observer must be prepared to provide a convincing reason for his presence in the community.
It may sometimes be advisable to let influential persons in the community handle the explanation of the investigator’s work. The observer then must decide upon the degree of his participation in the community, ranging from the bare minimum of answering when addressed, to engaging in some major activity concerning the community life.
[…] VIEW […]