09/11/2020

Research Methodology
Lecture 3: Hypothesis

Outline:
• Weekly plan
Identification Solution Implementation Validation Reporting
• Last lecture reminder
• Research questions
• Definition of a hypothesis © Prof. Wlodek Kulesza
• Types of hypotheses
1 2

1. Find a problem
2. Research the problem
Principles:
3. Form a research question
q 5 ECTS=>~135 hours +/- 5 hours
4. Form a hypothesis q All project activities can be counted in, including

5. Validate your hypothesis workshops (not seminars!).

* Plan and execute an experiment

* Analyze the results
6. Form a conclusion
7. Report
3 4

1
 09/11/2020

Choose a title that describes the thing you are

•Title of the project
investigating.
•Problem statement
The title should be short and summarize what the
•Research question(s) investigation will deal with. It should be a little
•Hypothesis exciting and promising.
•Main Contribution
•Keywords
•References

5 6

Write a statement that describes what you want to A very clear statement of the question is essential to proving
prove. that you have made an original and worthwhile contribution
to knowledge.
By making direct reference to your literature
review, you must demonstrate that your problem The question must be framed in terms of the variables you
• has not been previously answered, have identified. The question is asking about a relationship
between the variables.
• is worth answering.
Choose variables that you think act independently of each
e.g. as one can see from the reference list there is other.
still a luck of solution which would ….
e.g. ..the objective is to find out why (how) ….?

7 8

2
 09/11/2020

The examiners read your thesis to find the

When you know what variables may be involved you
answers to the following questions:
are ready to translate your questions into hypothesis.
• What is this student's research question?
• Is it a good question? A hypothesis is a question which has been
• Did the student convince me that the question was reworded into a form that can be tested by an
adequately answered? experiment.
• Has the student made an adequate contribution to
knowledge?

9 10

Make a list of your answers to the questions you

have. This can be a list of statements describing A hypothesis is a tentative answer to your research
how or why you think the observed things work. question.

There is usually one hypothesis for each question It tentatively explains certain behaviors, phenomena,
you have. or events that have occurred or will occur.

You must do at least one experiment to test each The hypothesis states researcher's expectations
hypothesis. concerning the relationship between the variables in
the research problem.

11 12

3
 09/11/2020

The hypothesis is formulated following the review of "A good hypothesis states as clearly and concisely as
related literature and prior to the execution of the possible the expected relationship (or difference)
study. between two variables and defines those variables
in operational, measurable terms."
The related literature leads the researcher to expect a
certain relationship. (Gay 1996, page 62).

13 14

Hypothesis can be classified in terms of how it is

To formulate a research hypothesis we start with a derived:
research question and generate operational
definitions for all variables, keeping in mind: • inductive hypothesis
- a generalization based on observations
• expected relationships or differences, - a theory building from case studies
• operational definitions.
• deductive hypothesis
- derived from theory

15 16

4
 09/11/2020

A directional hypothesis specifies not only that a

specific relationship exists, but, further, the direction Relational: I predict that, as excitement
of that relationship. increases, test performance will decrease.

For example: Causal: I predict that anxiety causes/leads to

poor performance on tests.
• non-directional: I predict that excitement and test
performance will be significantly related.

• directional: I predict that, as excitement increases, test

performance will decrease.
17 18

Related research questions Related hypotheses

Relational Causal Relational Causal

Non-directional Are stress and health related? Causal hypotheses must Non-directional I predict that stress and health Causal hypotheses must be
be directional. are related. directional.
Directional Are stress and health inversely Does increased levels of Directional I predict that stress and health I predict that increased
related, such that health stress lead to decreased are inversely related, such that levels of stress will lead
decreases as stress increases? levels of health? health decreases as stress to decreased levels of
increases. health.

19 20

5
 09/11/2020

A comparative hypothesis uses words such

• causal-comparative study as differ or greater/less than.
• correlational study Most experiments involve a comparative
hypothesis.

A correlational hypothesis talks about a

correlation between the two variables.

21 22

Hypotheses can be stated as: Statistical Hypotheses:

• Deterministic The statistical hypotheses consist of the null hypothesis

(H0) and the alternative hypothesis (H1).
• Statistical

23 24

6
 09/11/2020

Example:
If we let stand for the mean of the girls and stand
Research Hypothesis: for the mean of the boys.
Girls will achieve higher reading comprehension test
scores than boys at the end of the first grade. Our null hypothesis states that:
there is no difference between the two means on the
Operational Variables: reading comprehension test scores.
Reading comprehension will be measured by the Iowa
Tests of Educational Development, Reading The alternative hypothesis states that:
Comprehension, administered at the end of the year. the girls mean score on the reading comprehension
test will significantly exceed that of the boys.

25 26

The example:
A hypothesis for an experimental or causal-
comparative study An instructor investigates the relationship
between the number of minutes needed to
A hypothesis for correlational research complete an examination and the score on the
examination. He wants to use the data to
A hypothesis for descriptive research determine whether there is a significant negative
relationship between these two variables.

27 28

7
 09/11/2020

Example:
Research Hypothesis:
A researcher wants to find out the educational characteristics of
gifted students with learning disabilities in the middle school.
The length of time needed to complete an When asked what is meant by "educational characteristics" the
examination will be negatively correlated with researcher lists the following variables:
the score on the examination for college students.
- IQ,
- reading achievement,
- mathematics achievement,
- written language achievement,
- locus of control (internal versus external locus of control).

29 30

Operational Definitions:
Research Question:
• IQ - Scores on an individually administered IQ test
What is the average level of IQ, reading • Reading Comprehension - Scores on the Woodcock
comprehension, mathematics performance, Reading Mastery Tests
written language proficiency, and locus of control • Mathematics Performance - Scores on the Woodcock
for gifted children with learning disabilities in Johnson Tests of Achievement: Mathematics Section
• Written Language Proficiency - Scores on the Woodcock
the middle school.
Johnson Tests of Achievement: Written Language Section
• Locus of Control - A suitable measure for internal versus
external locus of control will be selected

31 32