Final Comments for Wendy



Is there anything that can’t be measured by psychologists?

In the field of psychology we often want to research abstract concepts which at first glance may seem impossible to measure or quantify. Often these concepts, which are referred to as hypothetical constructs are the central element of a research hypothesis and so in order for the research to be successful psychologists must find a way to measure this ( To achieve this, psychologist use operational definitions. These operational definitions are a collection of directly measurable behaviours or traits whose values can be combined to give information on the concept being looked at.


These operational definitions may be seen as the solution to a huge problem, however as always, there are pitfalls to be avoided in using them. The main problem is that you can use almost anything as an operational definition but it is not necessarily going to be valid. For example if you wanted to measure bravery, the amount of times somebody brushes their teeth a day would not be a good operational definition, however the willingness with which they leave their comfort zone could be (


However if we were not to use these operational definitions then we would have no way of measuring such variables. These operational definitions are the best method we have available to attempt to study in a more objective and standardized way. This allows for the same hypothetical construct to be investigated multiple times in a more reliable way.


It can be seen that with the use of operational definitions the things which can be measured by psychologists are almost infinite. However it should be remembered that when we use operational definitions we are not directly measuring the hypothetical constructs. Therefore there is certainly a vast amount of things that are simply impossible for psychologists to measure directly (happiness, bravery, resentment),  however we can still study them by measuring relating behaviours which are more overt and then applying this data to the construct.

The difference between a case study and single case designs

Case studies and single case studies, although sounding similar, are two very different methods which psychologists can use to collect their data. They vary in almost all aspects and are non-compatible to each other, that is that you could not expect to generalise the findings of one onto another.

By their nature case studies lend themselves to qualitative data collection. This is due to the fact that a large amount of information on a variety of mediums can be combined in a case study to get a general picture of the situation ( In a case study one individual is studied in great detail, with all information gathered relates to or originates from this person. Case studies are great tools for psychologists as the information gained is so extensive that they can use it in virtuously limitless ways. Examples of these methods used to great success include the H.M and Clive wearing case studies regarding memory. They have helped psychologists gain an applied insight into the way people use memory and what can happen when abnormalities occur. Such information would be impossible to gain with other methods which rely on more controlled conditions and the need for multiple participants.

However this benefit of case studies, the fact that such a wealth of information can be gained from one person is also its biggest weakness, a distinct lack of generalisation. It is impossible to generalise a case study back to the population because they are just one individual and the data gained is not derived from means, so it violates many of the assumptions required for quantitative comparisons ( They are also more subjective than other methods of research as they are more susceptible to the influence of the researcher’s emotions and interpretations.

Single case designs work in a different way. They are normally qualitative and involve the use of many participants. Each participant in a single case design acts as their own control meaning that there is reduced chance of the effect being down to individual differences ( These studies are repeatable as the method can be reapplied to other groups of participants. You can also generalise the single subject design back to the population as multiple means are compared. These studies lend themselves to exploration through statistical tests which make the findings easier to sum up.

It can be seen that these two methods though similar in name are very different once applied. It should also be noted that neither is better than the other as they both work better under different circumstances.

Is it good science to keep adding participants/manipulating data until you find an effect?

As scientists we invest a lot of time into collecting our data and finding ways to present it the world which might highlight its importance. However it is not uncommon to find a study which has manipulated its data or added participants in order to change the results and present an effect. This is contrary to the objective nature of science and therefore has no place within the scientific community.


Adding participants can help an experimenter to find an effect if their original participants did not show one. This can be used in combination with a measure of central tendency which is not particularly sensitive to the sample size, such as the median. ( This would allow an experimenter to present his findings as significant despite the fact that his original findings would not have shown this result. This would therefore limit the extent to which his results could be accurately related back to the general public, meaning they have validity issues which could be concealed.


Some experimenters add participants or manipulate the data because they have been studying something for a particularly long period of time and so have invested a lot of effort into it. They then become involved in the study to the point where they simply want to be proved right and so the scientific methodology of their study suffers. In this way an experimenter can appear to have proved their theory even though there was no real effect.


Although it can be argued that there must be something in the results in order to find an effect it should be noted that any real effect should be blatant enough to appear without the need to manipulate the data or call in extra participants. Mark twain popularized the saying about the three types of lies, “Lies, damned lies, and statistics.” This is not attitude scientists should want people to have towards statistics if we want them to believe our findings. The goal of science is to inform the masses, via the use of evidence. This evidence is most often in the form of some sort of statistics. Therefore it is good science to use statistics in an honest and informative way, any other use of them is bad practice and can only be detrimental to our field.

Comments for Wendy: WEEK 3


This one doesn’t link directly to my comment as I couldnt find an exact link I’m afraid. Hope you find the right comment 🙂