Chapter 2 Flashcards
How did Clever Hans give the appearance of answering questions, and how did Oskar
Pfungst unveil Hans’s methods?
Clever Hans has learned to distinguish and interpret subtle bodily movements of observers and used them
to comprehend expected (and often correct) answers and give those answers to fulfil the duty. Pfungst has
revealed Hans’s methods by conducting an experiment that involved manipulating Hans’s ability to
observe the people around him. By restricting Hans’s vision, having the person asking not know the
answer or removing the observers or the person asking from his sight, he found out that Hans couldn’t
interpret the answers when he didn’t have any subtle movements to use as a reference.
How are observations, theories, and hypotheses related to one another in scientific research?
The data provided by observations are interpreted by theories as theories use this data as a primary
basis. Hypotheses are specific, concrete predictions deducted from a vast and vague theory.
How does The Clever Hans story illustrate (1) the value of scepticism, (2) the value of
controlled experimentation, and (3) the need for researchers to avoid communicating their
expectations to subjects?
This story shows us, firstly, how valuable it is to not give in to the fascination of a strange claim and to
follow an objective path in questioning such claims. This way, the inquirer will lead a study to actually
investigate – and to disprove – these rather unbelievable claims, not just try to prove. Critical thinking
also plays a role in this situation, as Pfungst, according to his previous experience and knowledge about
the world, had to embrace a new point of view upon every failure to prove his claim. After formulating a
theory and hypothesizing, Pfungst also had to change his pattern of thought multiple times upon changing
variables in his experiment and gaining more results. He changed the ability of Hans to observe his
environment, the ability of inquirers to subtly move by seizing them the answers or moving them out of
Hans’s vision. He had recorded and interpreted these results, and sometimes advanced through his claims
as new information was obtained. He has changed the conditions in which Hans answered the questions
to gather new information – as well as to eliminate incorrect approaches. Also he had not let Hans
interpret his expectations as well by consciously ceasing his subtle movements. He has hid his
expectations while inquiring Hans’s abilities, and this has helped gain a clearer and more objective
perspective on what they depend on.
How can an experiment demonstrate the existence of a cause-effect relation between
two variables?
An experiment consists of different types of variables; namely the independent variable which is
hypothesized to affect the other variable when changed, and the dependent variable which is
hypothesized to be affected by the changes in the other variable. An experiment seeks to describe a
causal relation between these two by changing the independent variable and observing its effects on the
dependent variable. It is important to keep any other variable constant while conducting this study.
What were the independent and dependent variables in Pfungst’s experiment with Clever
Hans?
The independent variable was Hans’s ability to see, to determine if he would or would not be able to
obtain visual cues at a certain moment. The dependent variable was the percentage of questions Hans
responded correctly, defined by the state of the independent variable during each response.
What were the independent and dependent variables in DiMascio’s experiment on
treatments for depression? Why were the subjects randomly assigned to the different treatments
rather than allowed to choose their own treatment?
The independent variable was the type of treatment the subjects received – or hadn’t; the dependent
variable was their degree of depression after 16 weeks. The subjects were randomly assigned types of
treatment to eliminate any bias towards the results, as well as to balance the distribution of the
treatments fairly and equally.
What are the differences in procedure between a correlational study and an experiment?
How do the types of conclusions that can be drawn differ between a correlational study and an
experiment?
In experiments, the researchers have first-hand control over independent variable and can directly
observe what changes in an independent variable cause what differences in the dependent variable. In a
correlational study, however, for legal, practical or ethical reasons the researchers are not given direct
control over the situation. They observe various pre-existing conditions to assess a relation between
primal conditions and the outcomes. However, correlational studies do not define for certain a cause and
effect based relation between these variables as they cannot just be altered to serve the study.
How does an analysis of Baumrind’s classic study of parental disciplinary style illustrate
the difficulty of trying to infer cause and effect from a correlation?
It seems as the study establishes a direct relation between parental disciplinary styles and child behaviour;
but since the researcher did not interfere with any variable, it is highly possible – and likely – that there are
other aspects contributing to these classifications and their relations that were not taken into account.
Since in a correlational study it is not possible to manipulate certain conditions and observe their effects
on the outcome, any link that is being tried to be established between the disciplinary style and the
behaviour of children would be much weaker than it would be with an experiment.
How do descriptive studies differ, in method and purpose, from experiments and
from correlational studies?
Experiments and correlational studies both try to assess a relationship that varies in strength, but is
existent nonetheless. Descriptive studies, however, are only interested in the conditions and try to define a
certain behaviour, mind-set or a mental condition; their prevalence, increase etc. without trying to assert a
specific initiator or a set of conditions within a cause-effect spectrum.
What are the relative advantages and disadvantages of laboratory studies and field studies?
Laboratory studies offer greater control over the variables as it is a more controlled environment.
Researchers are able to manipulate the variables to their desire to observe clearly how the outcome
changes. However, the fact that the subject is under careful observation can sometimes alter their
behaviour and lead the researchers to record artificial, unnatural outcomes, which can then obscure the
results of the study. A field study is much more convenient in this case; the subject is in an environment that is a part of its everyday life, and so their observed behaviour is more natural and would probably
ensure a clearer outcome of the study. However the researchers’ control over the field is much scarce, so
it might not be possible to alter the variables to observe their effects clearly.
How do self-report methods, naturalistic observations and tests differ from one another?
What are some advantages and disadvantages of each?
Self-report methods are identified by the own statements of the subject about their behaviour and mental
state. Naturalistic observations, however, do not interfere with the subject or even come into contact –
they are mere observations of the subject in their ordinary life. In tests, the researcher deliberately
obstructs the subject’s pathway by presenting a problem or a question, and the factor of interference
separates this method from naturalistic observations. Self-report methods are a direct procedure to extract
exact answers from the subject without having to observe them and deduct possible explanations. While
being relatively fast and direct, self-report methods are often risky that the subject might –intentionally or
not – manipulate their answers and distort the outcome they may lead to. Also, self-report methods
depend on content that can be subjective, misinterpreted or remembered incorrectly; this fact renders
self-report methods’ reliability somewhat arguable. Naturalistic observations, on the other hand, present a
clearer image of what might a person’s everyday life, and the thoughts and emotions that they go through
in the meantime. This method ensures a more naturalistic answer, but due to the fact that the subjects are
aware of the researchers’ presence, they can –again, intentionally or not – change the way they behave
and reflect their thoughts and emotions, which can again obstruct a healthy analysis of the results. Tests
are convenient to perform and can be scored easily on a basic scale, however they are, by nature, artificial
and sometimes do not fully reflect how the behaviour, thoughts and emotions derived from the test results
could be applied to everyday situations.
How do the mean, median and standard deviation help describe a set of numbers?
The mean is the sum of the data divided by the number of data, the median is the score that is in the exact
middle when all the data is put in an order from the smallest to the largest, and the deviation is a measure
of how much the individual scores fall apart from the mean. These information help explain the
variability of a set of scores; which indicates how apart each data falls from each other and the mean.
How does a correlation coefficient describe the direction and strength of a correlation?
How can correlations be depicted in scatter plots?
A correlation coefficient calculation produces a score ranging from -1 to +1. The absolute value of the
score defines the strength of the correlation. Correlation between two variables can be placed onto this
scale. Results ranging from -1 to 0 show an inverse proportion between the variables which means an
increase in one of them can cause a decrease in the other. Results ranging from 0 to +1 describe a direct
proportion between the variables which means an increase in one results in an increase in the other, and a
decrease in one results in a decrease in the other. These correlations can be depicted on a scatter plot by
designing plots that show one of the variables and the one variable that is being tried to be measured, and
by pinpointing every variable within the plot to the exact point they fall on. If these points are tightly
clustered to form a line, it can be concluded that they have a strong correlation. If they are scattered across
the plot and do not form a visible line, the conclusion is that they have a weak correlation.
Why is it necessary to perform inferential statistics before drawing conclusions from the
data in a research study?
Inferential statistics are a method of assessing the real effect of the variables in a research study. It helps
understand how much of the results obtained were doings of “luck” – or uncontrollable facts, and how
much could actually be attributed to the variables that the researchers had accounted for.
How is statistical significance affected by the size of the effect, the number of subjects
or observations, and the variability of the scores within each group?
A large effect size is less likely to be caused by pure luck than a small effect size, therefore it is more
likely that large effects sizes constitute higher statistical significance. Similarly, larger subject numbers
render the chance of pure luck being the only factor significantly low. This means that the more subjects
there are, the more statistical significance it is possible to attribute to the study. The variability is the
spread of scores depending on variables accounted for or not, therefore less variability within subjects
indicates a lower effect of luck or chance, and a higher statistical significance of the data.
