Hypotheses
Adapted from Burns, R B (1997) Introduction to Research Methods, Longman
The planning stage is possibly the most demanding, certainly the most important part of the research process. If research were limited to gathering facts, knowledge could not advance.
Scientific method
proposal of a hypothesis
deduction from the hypothesis that certain phenomena should be observed in given circumstances
checking of this deduction by observation
Formulation
Hypotheses are the working instruments of theory
follows from the review of literature
problems are not the only source of hypotheses
they also derive from theory
a hypothesis must be stated so that it is can be either confirmed or refuted
Operational definition
use of indefinite terms in a hypothesis make it untestable. e.g. authoritarianism can only be tested if there is a scale of it
Variables must be defined because not everyone will have the same definition or understanding of the terms in a hypothesis and there are also differences in the precision. e.g. define Violence on TV
Research and Operational Hypotheses
the original hypothesis is often tightened up and refined and becomes the operational hypothesis.
Which of the following is testable?
Children who attend Sunday School show better acquaintance with the bible than those who don’t.
Children who attend Sunday School show greater moral fibre than those who don’t.
Judging hypotheses
concise, correct terminology
must be testable
must state the relationship between variables
hypotheses should be limited in scope, not universal.
they should be grounded in known facts
Hypothesis formation and testing
most research either looks for differences between groups or for relationships between groups
Hypothesis testing and significance
We are never proving a hypothesis, only testing it and eventually accepting ir or rejecting it
One tailed and two tailed hypotheses
One tailed makes its prediction in one direction, e.g. that something is more likely than not
Two tailed hypotheses are those that suggest that the effect of an independent variable could go either way. e.g. that something is different (without saying which way)
Type I and II Errors
A null hypothesis in a directional forms favours its rejection
If the null hypothesis is accepted but the null hypothesis is actually false, you have a Type 2 Error
If the null hypothesis is rejected but the null hypothesis is actually true, you have a Type 1 Error
Further reading:
Taylor, C. et al (1987) Thinking it Through: a practical guide to academic essay writing, Academic Skills Centre, Canada, pp.23-29
The Thesis Statement and p.26: exercises