FIC0154 Basic Comm Research
Reference materials: Lecture notes 4 and relevant text book chapters
______________________________________________
Question 1
Scientific research must be internally and externally valid. What do Internal Validity and External Validity mean?
Answers
Internal Validity
- Only investigate the topic that we set out to investigate, and not something else accidentally: results have to match the topic
Eg. Teenagers nowadays and their favourite pop stars: if question was asked after the death of that pop star, then the results are not accurate
- Ask questions that are not swaying the actual answers
Eg. Research on the rate of accidents: conduct during the rainy season - affects the results
- results shouldn't be biased
- researchers need to be aware of the several factors that could cause the results to change
External Validity
- can be generalized into a bigger population beyond current analysis
Eg. investigate some of the Malaysians: can also apply to the rest of the Malaysians
- research can be applied by others
- when conducting on a small group of people, we must ensure that the research can be applied to the society
Eg. Malaysians like nasi lemak or chicken rice? - conduct on 100 people (not all Malaysians) and then generalize: 40% of Malaysians like chicken rice and 60% like nasi lemak
- to have good generalization:
(a) sample must be random
(b) a mixed group of samples
(c) a representative for each group: to specify which group
Question 2
what do researchers need to take into account when ensuring that their research results are externally and internally valid?
Answers
(a) history
- events that can cause us to have internal validity
Eg. the death of Michael Jackson: results of research that was done before his death will be different from the one after his death
(b) maturation
- the biological deterioration (hunger, tiredness, sleepiness, etc) of the people you investigate on
Eg. typing speed: condition 1 - during daytime, condition 2 - during the night time after work
typing speed of respondent may be faster during daytime compared to those after work due to tiredness
(c) testing
- to see whether the procedure is done well and the instrument is complete
- to make sure results are not predicted or any procedures are skip that will affect the actual results
- researchers should do their experiment on different groups of people
(d) instrumentation
- also known as instrumentation decay
- the deterioration of instruments or methods
Eg. the recorder broke down: researchers are unable to listen to the response of the respondents
Eg. the person interviewing the respondents got tired
(e) experimental mortality
- subjects might drop out in the middle of the research: refuse to continue
- we cannot force someone to be in our research
Eg. 100 subjects: 60 drop out - results will be affected
- to avoid: get more subjects than needed
Eg. 100 subjects are needed but you get 120 subjects - in case any of them drop out in the middle of research
- getting subjects in the middle of research can also affect the results
Posted by: Foong Kah Mei (Clara) 0310003