Teachers’ Reference
Lesson 4 Experimental Designs
Learning Objectives
Students will:
have an understanding of the 3 types of causal experiments commonly used in scientific researches, Randomised Design, Prospective Design and Retrospective Design;
be able to identify the type of experimental design being used in the
establishment of causal links;
be able to know the pros and cons of the 3 types of causal experiments.
Suggested Lesson Time: 90 min.
Activity 1
A scientist, Dr. Chan, wants to design an experiment to investigate the following topic:
The effect of taking a nap (a short sleep during the day) on longevity
1. Please design an experiment to study this research topic using a control group and an experimental group.
(Use a flow chart / mind map to show your experimental design).
Note:
The subjects of this study / experiment will be selected and randomly divided into two groups prior to administering the suspected causal agent (sleep).
2. Please comment on the difficulties Dr Chan may face if he carries out his
research according to your design.
Expensive and time-consuming (for years) to carry out the experiment;
Must involve a large number of subjects for the experiment in order to have
statistically significant results
Must keep other variables constant (fair test) but in reality it is difficult to control.
Remarks:
If a scientist wants to study the effect of a certain poison on humans, he/she is not allowed to choose an experimental and control group at random from a large population.
3. What are the advantages of this kind of experimental design (Randomised
Experimental Design) in scientific investigation?
Randomised Experimental Design can provide strong evidence to support an hypothesis because other possible causal factors are controlled in the experimental design.
Activity 2:
Dr. Chan wanted to investigate the link between eating fruits and immunity from cold. He selected 5000 students randomly from 500 secondary schools in Hong Kong. Then he sent questionnaires to these students to collect information about their eating habits. He divided them into 2 groups: those who eat fruits regularly (eating fruits every day) and those who seldom eat fruits (eating fruits less than three times a week). The former became the experimental group and the latter became the control group. He found that the number of students in the experimental group outnumbered the number of students in the control group. He then pared down the size of the experimental group so that the numbers of students in the experimental and control group were the same.
He then sent another questionnaire to them to gather information from them on their frequency of getting cold in the following 6 months.
2. The experimental design in (1) is called Prospective Experimental Design.
What are its special features?
The subjects who have been exposed to the suspected cause (eating fruits regularly in (1)) are selected for the experimental group while the subjects of the control group who have not been exposed to the suspected cause (eating fruits less often) are selected for the control group.
3. What are the advantages of Prospective Experimental Design over Randomised Experimental Design?
Easier and less expensive to carry out the study/experiment;
May avoid ethical objections;
Can pool data from very large group of subjects, thus increasing the accuracy of the experimental result.
4. What are the limitations of this kind of experimental design?
The selection process of experimental and control group usually focuses on one single causal factor, hence it is possible that other factors will come into play at the very early stage of the selection process (i.e. the selected subjects having the suspected cause already affected by other factor(s)).
Activity 3
Another Design involves the collection of data prior to the setting of the objectives of the research and the design of the research. The data can be abstracted from past studies or researches. This kind of design is called Retrospective Design.
Retrospective Design: to find the Cause and Effect from past data and researches.
Key terms:
Suspected Cause (independent variable)
Suspected Effect (dependent variable)
Experimental group: a group of subjects chosen from the population with the suspected effect (or with improvement after treatment)
Control group: a group of subjects chosen from the population without the suspected effect (or without improvement after treatment)
Dr. X wants to use Retrospective Design to carry out a research on the relationship between acupuncture and the prolonging of life span of patients with last stage liver cancer. The following flow chart shows his experimental design:
Suspected Cause﹕acupuncture
Suspected effect: prolonging of life span of patients with last stage liver cancer
Task:
1. What would the result of Dr. X’s study be if acupuncture treatment can prolong the
life span of patients with last stage liver cancer?
We may look for the percentage difference (life span) between the experimental and control group. If there is a higher and significant percentage of patients who can still live after 1 year in the experimental group than that of the control group, the study may support the causal link between acupuncture and prolonging of life span of patients with last stage liver cancer.
2. What are the pros and cons of Retrospective Experimental Design?
The advantages of retrospective experimental design are that it is quick and inexpensive. We only need to analyse the data of past research/study carefully.
The disadvantages of retrospective experimental design are as follows:
It can only provide weak evidence for a causal link because there may be other
potential causal factors which are difficult to control in this kind of experimental
design (these potential causal factors other than the one to be tested for may
automatically be built in the experimental and control group).
Furthermore, retrospective experimental design cannot estimate the level of difference of the effect being studied.
Reference: Stephen S. Carey (1998)
Exercise:
Read the following news and answer the questions:
Inhalation of air with high oxygen concentration enhances memory
A Japanese electric appliance Company X conducted a research and claimed that inhalation of air with high oxygen concentration can enhance memory. 80 students were divided into 2 groups. They were required to take a test on English vocabulary. Then the experimental group of students inhaled air of high concentration of oxygen while the control group breathed ordinary air. Both groups of students were allowed to have some revision before taking the same vocabulary test again. It was found that the experimental group could memorise the vocabulary items 15% more than that of the control group.
1. Name the experimental design used by Company X.
2. Comment on the experimental result and judge the reliability of the research
findings.