Research Methods

...

hypothesis→ often reverse of what researcher actually believes

- Alternate hypothesis→ what researcher believes

- Try to falsify null hypothesis → easier to show something is wrong than to prove something is right

- If the null hypothesis isn’t true they assume some increased support for the alternate hypothesis

3) DEFINING HOW ONSTRUCTS WILL BE MEASURED

- Better defined our construct the better our chances of finding or developing a good measure

- Constructs are however difficult to measure because the empirical representation must capture several elements in the definition

SAMPLING

- representative sampling→ the process of samling a population in such a way that the data from the sample is representative of the larger population

- sample selection must be unbiased→ minimize sampling bias through random selection of sample. Everyone in population has equal probability of being chosen

- Must have a significantly large sample size

RESEARCH DESIGN

1) Experiments → in the lab in the field

2) Surveys (non experiment)

3) Observational research (non experiment)

DETERMINE APPROPRIATE STUDY DESIGN

1) EXPERIMENTAL → participants are randomly assigned to one of a number of ‘conditions’ (situations e.g. participants carry out their job or do something on computer). There are usually one or more intervention groups (treating participants in a way that reflects some aspect of our theory) and a control groups (not treating them in accordance with our theory).

2) QUASI EXPERIMENTAL→ Participants are not randomly assigned to conditions. Usually participants are already in naturally occurring groups such as pre existing work teams, shifts or production areas. Problem with this is that not all control variables are controlled e.g. different shifts may face different conditions therefore we cannot necessarily say that the independent variable changes dependent

3) NON EXPERIMENTAL→ survey research, but control variables are not controlled. Observational research (researcher acts as a fly on the wall)→ case studies/ useful for exploratory research and discovering new ideas

ISSUES IN ORGANISATIONAL RESEARCH

- MULTIPLE CAUSATION→ more than one variable affects/ humans are complex

ETHICS IN ORGANISATIONAL RESEARCH

- Do not harm

- Treat ppl with dignity and respect

- Ensure privacy and confidentiality

- Participants must be well informed about the research and the procedures

- Their willing consent to participate must be obtained

- They must be free to stop participating at any time