Placebo and experimenter effect

advertisement
Placebo and experimenter effect:
Placebo effect- refers to the participants behavioural response being
influenced by their expectation of how they should behave. The placebo
effect means that the expectations of the participants, rather than (or as well
as) the independent variable, may be affecting the dependent variable, and
therefore the results of the experiment. The placebo effect can be minimised
by using a single-blind procedure in allocating participants to groups. The
single-blind procedure is the term given to the method of allocating
participants to groups so that they are ‘blind’ to the knowledge of which
group they are in: experimental or control.
Experimenter effect- the experimenter can also influence the outcome of
the experiment by unintentionally influencing the behaviour of the
participants known to be in either the control or experimental groups. The
experimenter effect refers to the actions of the experimenter, rather than
(or as well as) the independent variable, affecting the dependent variable,
and therefore the results of the experiment.
One type of experimenter effect is known as the self-fulfilling prophecy.
The self-fulfilling prophecy is a term used to describe a situation in which
participants behave in line with the way they believe the experimenter wants
them to behave. Subtle differences in facial expressions (such as smiling
when delivering the instructions to one group, but not to the other) can bring
about a change in the expectations of participants in the respective groups.
Another type of experimenter effect is referred to as experimenter bias.
Experimenter bias refers to the unintentional actions that may occur in the
collection and treatment of experimental data. It is most likely to occur when
the experimenter wants to find a certain pattern of results to support the
hypothesis being tested. It includes a range of behaviours such as incorrectly
reading raw data, wrongly interpreting a participant’s response, or even
unintentionally giving assistance to participants in one group and not the
other.
The experimenter effect can be minimised by using a double-blind procedure
for the allocation of participants to groups.
The double-blind procedure is the term given to the method of allocating
participants to groups so that both the experimenter and the participants are
“blind” to the conditions to which the participants have been allocated. This
procedure should be used whenever there is a risk that the experimenter and
the participants both might influence the results by their own expectations
and desires.
Minimising Effects Of Extraneous Variables
To minimise the effects of extraneous variables, experimenters try to ensure
that the groups being compared (experimental and control) are as similar as
possible with respect to the distribution of relevant personal characteristics.
This is done by controlling the manner in which participants are allocated to
groups; part of the experimental design. Three experimental designs are:
1. Repeated Measures Design- sees each participant involved in both
the experimental and the control conditions. This design controls or
eliminates any effects that might be attributed to the personal
characteristics of the participants, since they remain constant. Also
known as the within participants design, the repeated measures
procedure can suffer from another extraneous variable; the order
effect. Participants may perform better on the task when doing it a
second time because of the effect of practice. Conversely, they may do
worse the second time because of fatigue. One way of minimising the
order effect is to space the time between the two conditions.
Participants may be given a week or more between performing under
the experimental and then the control condition. However, a safer
practice is to counterbalance. Counterbalancing involves half the
participants performing the task under experimental conditions first
and control conditions second and the other half of the participants
performing the task in reverse order to this. Even the decision as to
which participant perform under the experimental condition first can be
done by random selection.
2. Matched participants design- involves selecting participants on the
basis of scores achieved on a pre-test or a number of pre-tests. In this
way, the participants can be ranked in accordance with their scores
and then allocated to the respective groups. For example, suppose
participants were pre-tested for reaction time before undertaking an
experiment that hypothesised that alcohol impaired reaction time. The
two participants with the quickest reactions on the pre-test would be
randomly allocated; one to the experimental group and one to the
control group. Then the third and fourth fastest would be allocated in
the same way until all participants had been allocated to one group or
the other. In such an experiment, it would also be wise to take
personal attributes such as sex and weight into account. If the groups
happened to end up with a high proportion of large males in the
control group and a high proportion of small females in the
experimental group, the difference in metabolism of alcohol due to the
difference in build could also become an extraneous variable.
3. Independent Groups Design- allocated participants to groups via a
random procedure such as the toss of a coin, or by having each
participant draw a numbered ticket from a container that holds the
same number of tickets as participants. Then, all those who draw an
even number might be allocated to the control group and all those who
draw an odd number to the experimental group. The independent
groups design is also referred to as the between participants design.
Download