It was only the mental activities demanding the use of the head that attracted attention. It is quite pleasing to note that with time evaluation was extended to the affective dealing with feelings.
The observation technique, as the name indicates, is a way of observing a person as he exhibits certain selected behavioral traits. It may make use of designed instruments or follow reports of all that happened during the period of observation.
The reliability of the observation technique may be hampered by subjectivity in the reporting by different observers of the same event.
To minimize disparity in reporting observed traits a systematic – observation instrument may be used. The observer awards tally for predetermined traits as they occur. The expected behaviors may be pre-coded or coded immediately after the observations.
The recordings are then processed. In recent times, a video – recording of events is made used. The camera should be focused on only traits that have been earmarked for observation otherwise unintended measures may obscure the ones scheduled for observation.
At the end of the exercise, the video recording is edited and the recorded behaviors are coded. The behaviors to be observed must be listed and well defined so that an observer can say with preciseness the occurrence or non – occurrence.
The method of the recording should be stated without any ambiguity. The recording makes use of two systems.
Sign System: In the sign system a list of attributes ranging in number from sixty to one hundred to be observed is contained on the recording sheet. Each time an attribute is observed a tick is marked in front of that attribute is a tick is marked in front of that attribute.
The observer is expected to mark as many attributes as he observes during the period of observation. Multiple marking for an attribute is disallowed no matter how often the attribute is observed. The focus is wide but inactive.
Category System: This contains a limited number of attributes to be observed. A tally is placed in front of each attribute each time it is observed. The focus is limited but active.
For any system to be adopted, it is better for the designer to explicitly define the behaviors to be used. Some form of training is essential for the use of the instruments.
There should be inter and intra-correlational study of the instrument on a video-recorded event. With the training on the use and comparative study of results obtained by the same at different times or different people at a time, the reliability and validity of the instrument may be guaranteed.
A questionnaire contains a set of items that require the supply of information on certain issues by a respondent. The questions may be open-ended or restricted.
The open-ended questions allow the respondent opportunity to respond to questions freely while restricted questions limit the choice of answer to one or more of the given options. Restricted questions may provide a checklist or rating scales for the use of the respondent.
The checklist is to be ticked at any point considered appropriate just as the rating scale provides a continuum along which the respondent indicates what he considers appropriate.
Questionnaires are in some cases preferable to interviews as they are less expensive and easy to administer, easy to mail to the respondents, and can be administered to a large group of people at a time.
Do not necessarily involve interpersonal action between the respondents and the executor of the questionnaire, contains no restriction on a time limit for its completion, makes responses anonymous and exposes all respondents to the same questions.
Questionnaires, like any instrument, have their limitations. The completion of questionnaires may lack sincerity and be deficient in accuracy.
Errors of respondents need to save themselves from embarrassment, and misinterpretation of the questionnaire as some people see no reason why they should complete a questionnaire for a researcher to earn a degree while they earn nothing but spend their time.
Faking in response may be easily identified if the questionnaire contains similar questions put in different ways but this may not be so for all issues touched upon by the questionnaire.
Mailed questionnaires often suffer a high attrition rate in returns of completed questionnaires. The low or late return may invalidate whatever results are obtained. Administration of questionnaires to a large group of people who will fill them on the spot the moment and return immediately is much better.
To enjoy a high rate of returns and somewhat reliable responses the purpose of the questionnaire must be disclosed to the respondent and the sponsor named. The questions must be brief and the respondents assured of anonymity of their responses.
The time of administering a questionnaire, the mood of the respondents at the time of completing the questionnaire, and the environment of the respondent are all capable of influencing the information supplied on a questionnaire.
We, therefore, say that the standardization of both the questionnaire and the testing situation functions reliability of responses. The construction of questionnaires should be guided by:
- Decision as to what information are desirable
- Decision on whether to use structured or unstructured items.
The structured form is to be preferred for ease of analysis.
- The draft of items should be trial tested and necessary modifications made before arriving at a final one.
- Questions should be brief and precise.
- Precode the items so that analysis can be done for a large group using the computer no sooner than the test is administered.
Ratings: Assessment may at times be made based on personal impressions. The classification to be used may take such forms as:
(i) Excellent, Very Good, Good, Poor, Very Poor.
(ii) Outstanding, above average, average, below average, unsatisfactory.
(iii) Strongly agree, agree, undecided, disagree, strongly disagree.
It assigns numbers with specified values on a defined continuum. Both ends of the continuum must be well defined.
Rating scales do not make use of raw scores but assign some subjective values to behaviors or expressed opinions.
Ratings are of limited value as they are easily rendered unreliable by such factors as:
- Raters bias: A rater is not likely to give a correct rating to someone who is his friend or relation.
- Rater’s guess: A rater may use quick guessing to rate his respondent. This may happen when there is a time constraint.
- Jumping at conclusions: The impatience of some raters to follow through with the item to be rated before carrying out the rating.
- First impression syndrome: Some raters are so emotional that they are easily carried away by the first impression. The later work is never critically examined, the earlier part determines the rating.
- Appearance: The appearance of an individual rather than the actual response may in some cases dictate the rating.
- Prejudices: Raters with sexual, tribal, or racial bias may prove very unreliable in rating.
- Halo effect: The personal feeling of a rater may influence his rating.
- Central Tendency Error: Some raters prefer to stand on the fence. They want to be seen as neither strict nor generous, they, therefore, rate most work as average.
- Leniency: A rater may not want anybody to score low even though the responses deserve low scores and so may rate poor responses as average.
Scales construction must be a joint venture of both the assessor and the assessee. They should both agree on the items as well as the procedure for carrying out the rating.
It is in line with this that the Annual Performance Evaluation report was prepared and the assessment has to be discussed by the assessor with the assessee who passes his comment on whether he agrees or disagrees with his reporting officer’s scoring.