Vous êtes sur la page 1sur 6

Ms Nicola Erasmus

W2065466
ED209 TMA 02
24th April 2009
Word Count: 1998

Part 1 Short-answer questions


a) Correlation co-efficient
Correlation, in the statistical sense, refers to a relationship between two or more
variables. As one variable changes, the other variable changes.

For example, one variable might be the number of hunters in a region and the other
variable could be the deer population. Perhaps as the number of hunters increases,
the deer population decreases. This is an example of a negative correlation (a – sign is
in front of the correlation co-efficient). A positive correlation (a + sign is in front of the
correlation co-efficient) is where the two variables react in the same way, increasing
or decreasing together. For example, temperature in Celsius and Fahrenheit have a
positive correlation.

The correlation coefficient, denoted by r, is a measure of the strength of the linear


relationship between two variables. The correlation coefficient takes on values ranging
between +1 and -1. Thus, a correlation co-efficient of 0.7, would indicate a strong
positive relationship between two variables.

b) Between-group comparison participant allocation

In research, statistical tests are performed to determine if the difference between


groups is greater than can be attributed to chance. An experimental design study
tests whether there is a causal relationship between two variables. A between-group
comparison involves statistical comparison of the outcomes of one group with another
(not changes within a group). One group experiences a specific condition and the
other group (control group) does not. In all other respects, the groups are treated the
same. If differences are observed between these two groups, then the differences
inferred are due to the fact that one experienced the condition and the other did not.
The researcher chooses two groups to make these comparisons. The allocation to
these groups may either be done randomly, for example: flipping of a coin – ‘heads’ =
group A and ‘tails’ = group B. However, the participants are from a common pool, but
the groups created cannot be stated to be exactly the same since they are randomly
assigned to the group. The main disadvantage of random selection of participants is
that by chance, all participants with one characteristic may be allocated to one group.
This may confound the results. The main advantage is that the participants are less
likely to work out the purpose of the study.
Another manner of allocation of participants to the groups is by matching the
participants in terms of certain characteristics that are relevant to the research. For
example, age and gender. Participants are less likely to become bored or tired with
the study.
c) Time Sampling
Time sampling is used to study and observe behaviours. Predetermined units of time
(ranging between one and sixty seconds) are used to guide the observer’s attention
Ms Nicola Erasmus
W2065466
ED209 TMA 02
24th April 2009
Word Count: 1998

throughout the observational period. The length of the time unit is dependent on the
frequency of the behaviour. For example, observing blinking would require shorter
time intervals. This observed behaviour is then coded into specific categories, for
example, interaction between family members.

The two types of time sampling techniques include one zero and instantaneous
coding. Within instantaneous time sampling, observation and coding are done at the
same time. For example, at the 5 second segment, a child interacting with a family
member is categorised. On the other hand, one zero time sampling technique uses a
determined time segment and the observer stipulates if a particular behaviour is
present (one ) or not present (zero) and records before moving on to the next subject.
For example, during the 5 second time segment, interaction between family members
being present or not is categorised.

During the recording, the observer uses a recording grid to collect information in clear
sets of defined categories. If working in the ‘field’, a stopwatch or buzzer may be used
to signal to the observer the end of the allocated time unit for observation.
Specialised computer software programmes are also available to aid in time sampling
and coding of a video of the behaviour exhibited over a time period.

Behaviour may be missed between these time samples, but it is thought that over
time it will produce an accurate picture of the behaviour occurrence.

d) Quasi-experimental Design
A true experimental design includes random assignment of participants to groups and
manipulates the independent variable. Therefore the outcome on the dependent
variable can be examined. Often in psychology we may want to look at variables
which we cannot manipulate. If a comparison needed to be made between males and
females, random allocation of participants would not be possible; since the
participants are either male or female.
A quasi-experimental design lacks random assignment of participants to groups.
Without this random assignment of participants to groups and manipulation of the
independent variable, a researcher is not able to infer cause and effect conclusions.
Quasi-experiments do not meet the requirements necessary to control extraneous
variables.
Ms Nicola Erasmus
W2065466
ED209 TMA 02
24th April 2009
Word Count: 1998

An example of a quasi experimental design might be a study examining the effects of


smoking on respiratory functioning. However, participants may smoke one pack a day
while others may smoke two packs a day. Yet, it would not be ethical to expect the one
pack a day smokers to smoke two packs. The study conclusion would not be able to
make any cause and effect conclusions.

e) Reliable observations
A test is considered reliable if we repeatedly get the same result - results are
consistent. For example, if a test is designed to measure a trait (such as introversion),
then each time the test is administered to a subject, the results should be
approximately the same.
When observing and coding a child’s behaviour, objectivity and a detached disposition
are vital. However, this skill would need to be practiced so as to ensure that reliable
observations are made. Intra-observer reliability could be obtained by administering
the test twice at different points in time.
One also needs to identify and clearly define what is being observed and how it is
being observed. This in turn will prompt an operational definition with a well defined
coding system. Another manner to test reliability would be to get another coder to
code the behaviour and compare the results (inter-reliability).

Part 2 Ethics and Research Design


a) Compare types of comparisons made in original experiment and
replication.

Reviewing the experimental design between Bandura’s original experiment (1965)


and the BBC’s replication (2000); Bandura utilised the between-group design, whereas
the replication experiment used the within-group comparison design ( FOCUS II and
Neural Networks, 2006).

Bandura’s between-group design grouped 4yrs olds into 3 different groups. Each
group watched a video of a man acting aggressively towards a doll (Bobo doll). Yet,
each group experienced a different ending to the video i.e. punishment to the man for
acting aggressively, praise for the man for acting aggressively and no consequence
for the man for the aggressive behaviour. This video watching would be termed the
independent variable. The children of all three groups were observed in a room with
the same Bobo doll after watching the video. This data collection showed that the
dependent variable of aggressive behaviour mimicked by the children correlated with
the hypothesis that children’s behaviour is affected by television (FOCUS II and Neural
Networks, 2006).
Ms Nicola Erasmus
W2065466
ED209 TMA 02
24th April 2009
Word Count: 1998

However, the BBC replication experiment used the within-group comparison design
study. All participants (aged 3yrs) experienced the same conditions (independent
variable). Each child watched the same video of a man acting gently towards the Bobo
doll and in a separate video, watched a man acting aggressively towards the Bobo
doll. The observed children’s behaviour showed that 2 of the three participants
mimicked the kind behaviour, whereas all of the participants mimicked the aggressive
behaviour to varying degrees. This study did not incorporate the system of reward or
punishment for behaviour. Thus, the hypothesis of children learning behaviour from
television was also supported by the data collection in this study (Focus II and Neural
Networks, 2006).

b) Ethical issues raised by the replication study


With reference to the British Psychological Society’s principles for conducting research
with human participants (Method and Skills Handbook,2006) , a few ethical issues of
concern may be raised regarding the replication of Bandura’s (1965) Bobo doll study
as shown on the Focus II and Neural Networks DVD, 2006 – investigating if what
children watch on television influences their behaviour.

The replication research study may be criticised for aggressive behaviour shown to
the children via the television. All three children copied this aggressive behaviour, to a
degree, when taken into a room with the same Bobo doll. This aggressive behaviour
could result in a permanent behaviour change in the children. An example of this may
be, Helena. Although she gave a gentle hit to the doll, she is not usually aggressive at
all. William may also be another example, as he may usually play in a rough manner
with his sibling (as his mother describes). Yet, he copied the aggressive behaviour
towards the doll for six minutes and thirty eight seconds. William would not even “kiss
the doll better” (Focus II and Neural Networks, 2006) as requested by his mother.
William had initially been two of the three children to have imitated the kiss towards
the doll from the initial video clip (Focus II and Neural Networks, 2006). Thus, the
ethical concern in these examples used would be that there may indeed be
Ms Nicola Erasmus
W2065466
ED209 TMA 02
24th April 2009
Word Count: 1998

“...negative consequences for the individual’s future life...” (Methods and Skills
Handbook, 2006).

The ethical concern of protection of participants may also be raised in the Bobo Doll
replication study, that is, “...protect participants from physical and mental harm
during the investigation” (Focus II and Neural Networks, 2006). The aggression shown
to the children, resulting in them copying the aggressive behaviour towards the Bobo
doll, may have caused the children personal distress, which could have a negative
effect on their lives.An example of this may be Helena. Helena was the only female
participant in the replication study. She is viewed as not being a likely candidate to
show aggression towards the doll. Even she “...gives a few lady-like hits with the
mallet...” (Focus II and Neural Networks, 2006).

The children were given a colouring crayon with a piece of paper while watching the
video of the Bobo Doll replication study. They also wore sensors on the fingers of one
of their hands while watching the video clip (Focus II and Neural Networks, 2006).
This inferred that perhaps they were deceived regarding the purpose and nature of
the experiment. Although consent would have been obtained from their parent, due to
their age and the participation of their parent, the children may have felt pressurised
“...to take part in, or remain in, an investigation (Methods and Skills Handbook, 2006).
The children may have viewed it as a game and not understood it as being a study
and observation of their behaviour in the play room.

c) Address these ethical issues if running a similar replication study

In terms of addressing the ethical issue of perhaps permanent aggressive behaviour


change occurring in the children, one would need to consider the debriefing process
carefully. “Debriefing does not provide justification for unethical aspects of any
investigation” (Methods and Skills Handbook, 2006). However, in this case, the ethical
issues may be avoided if the nature of the study was explained individually, at their
linguistic and cognitive level. Two debriefing sessions may be carried out with the
participants. Initially, an explanatory session regarding the study. The other session
could include an active intervention session, in which both the investigator and the
parent talk to the child - clarifying that it was a study and that the aggressive
behaviour is an unacceptable (re-iterating acceptable behaviour outside of a
laboratory setting). These debriefing sessions would need to be completed before the
participants leave the study environment and then again after a specific time period
to note any behavioural changes within the home environment and deal with these in
an appropriate manner if required.

Regarding participants feeling pressurised into participation and remaining in the


investigation, perhaps a buzzer type watch could be placed on the child’s wrist. If at
any time the child felt uncomfortable, he could buzz the investigator to leave and
return to a parent. This might then re-assure the child that their parent is available to
help them if they experienced any form of distress, thus eliminating the feeling of
Ms Nicola Erasmus
W2065466
ED209 TMA 02
24th April 2009
Word Count: 1998

participation pressure. A debriefing session should once again be held with this
participant and parent to acknowledge and deal with any issues which may have
arisen for the child.

References

FOCUS II and Neural Networks (2006) ED209 Child Development. [CD-ROM]


Milton Keynes, The Open University.

Oates, Sheehy and Wood (2006) ‘Theories of Development’, in Oates, J., Wood, C., and
Grayson, A. (2006) Psychological Development and Early Childhood, Oxford, Blackwell/
The Open University.
The Open University (2006) Media Kit, Band (Zero to Hero)
ED209 Child Development. [CD-ROM] Milton Keynes, The Open University.

Vous aimerez peut-être aussi