ADVERTISEMENTS:
R.L. Ackoff has attempted to answer this question in a systematic way. All research problems ultimately reduce to the question, which of a set of alternative means is the most efficient one. Once those alternative means are formulated, the researcher is in a position to pose a question of each of the means, as to what may constitute the evidence that this particular means is the most efficient one among the alternatives.
The answer to this questions would usually be in the form:
“The particular means can be accepted as the most efficient among the alternatives under specific conditions.” Such specific conditions should be formulated for each of the alternative means. The statements of these acceptance conditions are the hypotheses. The researcher does not, of course, know which of these alternative hypotheses is true; this is precisely what the research is designed to determine.
ADVERTISEMENTS:
Ideally, a researcher should start with trying to determine all the alternative means (solutions of explanations) of coming to grips with his problem. This means that the researcher needs to undertake a ‘resource survey’ which includes survey of related theories or orientations, which may bring to light what alternative means, solutions or explanations may be applied to the problem.
The researcher will attempt to determine which of the alternative course of action or solution or explanation is most efficient in terms of certain criteria, e.g., economy predictability etc. Let us now suppose that a researcher has a problem whose solution depends on certain predictions and the researcher knows that there are three alternative theories (means) which are germane to the problem.
Now, if one of the three theories is more likely to predict events more accurately than the other two, it may be taken as the most efficient one as a solution to the problem. If the problem happens to be one dealing with practical or programmatic concerns, the criterion of efficiency of alternative course of action may be economy in the realms, of time, money and energy.
The alternative hypotheses which the researcher sets out to formulate are nothing but statement of conditions for each of the alternative means under which conditions, it (each alternative means) may be seen to be the most efficient.
ADVERTISEMENTS:
Simply stated, the alternative hypotheses are the statements of acceptance conditions for each of the alternative courses of action or alternative solutions to the problem. Suppose the researcher’s problem is to decide which of the two types of teaching methods should be recommended for a particular educational institution.
The research decides to use, let us say, the student’s examination scores as a measure of efficiency (of the means).
Then, for each of the alternative teaching methods, his statement of acceptance conditions, i.e., alternative hypotheses, will be as under:
H1:
The average examination score produced by teaching method No. 1 is greater than the average test score produced by teaching method No. 2.
H2:
The average examination score produced by teaching method No. 2 is greater than that of teaching method No. 1. Therefore recommend Method No. 2 if H2 proves to be correct.
We note here that one possible outcome has not been considered, i.e., the test scores are equal for both teaching methods (No. 1 and No. 2). Now, if the test scores were really equal, i.e., if both methods were equally efficient, the researcher will have no course of action to select for recommendation, consequently, he may have to add another course of action.
It is clear now that formulation of alternative hypotheses involves the following steps:
ADVERTISEMENTS:
(1) A measure of efficiency applicable to all the alternative courses of action is selected (Examination Score: sales, productivity etc.)
(2) On the basis of this selected measure of efficiency a set of acceptance conditions for each alternative course of action is assigned.
(3) The acceptance conditions are reformulated as hypotheses which are mutually exclusive and jointly exhaustive.
In all research (theoretical or action oriented) alternative courses of action (solutions, explanations) acceptance conditions (economy, predictions etc.) or hypotheses should be made explicit.
In fact, if the acceptance of one set of alternative hypotheses rather than another would make no difference whatsoever to subsequent behaviour (scientific or practical) then the problem or its formulation is scientifically meaningless.
It is obvious that there is no scientific way of selecting one of the alternative hypotheses as valid save when there is some index of efficiency which can be applied to each of the alternative courses of action. The applicability of the measure of efficiency to alternative course of action depends on certain conditions holding.
For example, in our illustration of the alternative methods of teaching, the use of examination score as a measure of efficiency may be suitable only if each student is allowed an equal period to complete the common test.
Such conditions constitute the points of agreement among the hypotheses. These points of agreement among the hypotheses are either known or assumed to be valid. Should such an assumption be made, the researcher must make it explicit.
If the researcher sets up two hypotheses, there must be at least one point of agreement among them and one point of variance or disagreement.
These alternative hypotheses may be represented symbolically as under:
H1 – MN1
H2 – MN2
H3 – MN3
H4 – MN4
The alternative hypotheses should encompass all possible outcomes of research, that is they should be exhaustive with respect to the points of disagreement which will be tested. Secondly, of course, the hypotheses should be mutually exclusive.
Failing these two requirements, research will not indicate which one course of action or solution should be selected from among the entire range of possibilities represented by the alternative hypotheses.
A very effective way of assuring ourselves that the hypotheses are mutually exclusive and jointly exhaustive of the universe of possibilities is to use the logical technique known as the “Boolean expansion.”
Suppose we have one common point of agreement (M) among the later native hypotheses and three points of disagreement (e.g., N, O and P), then the alternative hypotheses according to the requirements of exhaustiveness and mutual exclusion could be presented as under.
The common point of agreement among these hypotheses may be the examination score under specific conditions. Thus, M = Examination score. The points of difference may be N = more than x; N’ = less than x; similarly O = more than y, O’ = less than y and P = more than z, p’ = less than z.
(Read H4 as Examination score is more than x and y but less than z.)
In general, if there are in points of disagreement, there will be 2 n (2 x 2 x 2 x 2.n times) alternative hypotheses in an exclusive classification. Only one of them can be true and must be true.
In a research involving more than two hypotheses, it is advisable to formulate the points of disagreement symbolically in a manner indicated to facilitate the construction of hypotheses. Intuition is often not a satisfactory guide.
It was suggested earlier that ideally there should be one hypothesis for each alternative course of action. Such a problem is one which involves estimation, e.g., estimating an optimum number of workers for a production unit 100, 250, 300 etc.
The selection of the most efficient course of action depends on an estimate of the value of a critical variable (i.e., the exact number of workers) in such cases, it is not economical to formulate explicitly each alterative course of action and to associate a hypothesis with respect to each. We can use only a shorthand formulation.
The alternate hypothesis can simply be stated as: “K workers are needed” and the problem of research is to estimate this K. Now, since the estimate of the value of any variable is subject to error, it is advisable to express the estimate as a range of values rather than a single value, e.g., 300 = 50 workers (250 to 350) are needed.