2. Gives direction to the collection and interpretation of the data will guide the research design: variables, sampling, time frame, instruments, statistical methods, etc.

Five principal purposes served by the hypothesis

1. Provides a tentative explanation of phenomena

2. Facilitates the extension of knowledge in an area

3. Provides a relational statement that is directly testable in a research study -- focus on relationships that are amenable to direct observation

4. Provides direction to the study -- keeps it restricted in scope; identifies the parameters for the study

5. Provides a framework for reporting the conclusions of the study

Two types of research hypotheses

1. Inductive Hypothesis

Formulated as a generalization from observed relationships

Researcher makes observations of behavior

Notices trends or probable relationships

Hypothesizes an explanation for this observed behavior

Must be accompanied by an examination of previous research

Put forth as a suggested solution / answer to research questions based on identification of a problem

2. Deductive Hypothesis

Derived by deduction from theory

Predictions for research questions grounded in theory

Advantage - leads to a more general system of knowledge, because the framework for incorporating them meaningfully into the body of knowledge already exists within the theory itself

Theories are general explanations that apply to a wide range of phenomena; From the inter-relationships proposed in the theory, one
can state specific consequences that could logically be assumed to follow. These consequences form the basis for hypotheses and must be checked for any logical gaps between theory and hypothesis.

Five criteria of a useable hypothesis

1. Must have explanatory power

2. Must state the expected relationship between variables (directional versus non-directional) that specifies the direction of the relationship versus not predicting the direction

3. Must be testable

4. Should be consistent with the existing body of knowledge (another reason for comprehensive review of the literature)

5. Should be stated as simply and concisely as possible (principal of parsimony -- simpler is better)

Evaluation of Hypothesis

1) does it express a measurable relationship between and among variables?

2) is it written as a declarative statement?

3) is it related to a theoretical statement?

4) does it use statistical terms?

5) is it written in terms of relationships between constructs?

6) does it use words which indicate a value judgement?

"More hypotheses are rejected than are supported". (Darwin)

Characteristics of Null (or Statistical ) Hypothesis

stated in form of no impact or no effect or no relationship

is the only hypothesis tested by statistical analysis (against chance)

used because it enables researchers to compare their findings with chance expectations through statistical tests

assumed to be true until there is sufficient evidence to indicate otherwise

assumes that any differences occurred due to chance

determines the probability that the null hypothesis is true

If the observed differences could easily be a function of chance, the evidence is insufficient and we fail to reject the null hypothesis.

If the observed differences could not easily be a function of chance, the evidence is sufficient to reject the null hypothesis in favor of an alternative hypothesis.

if found to be rejected, then research hypothesis is supported

if fail to reject null hypothesis, then theory will need revision

can not prove

leads to future areas of research

Readings