Archive | October 2011

How do politics affect the science of psychology?

Politics can affect the science of psychology in many ways and I do believe that politics have a great impact on psychology. To begin with political correctness has come to the point where it is laughable for example pigs were removed from a children’s toy farm set in case having them in the set would offend Jewish and Muslim parents, the whole story is on . What my point is, how many Jewish or Muslim parents would open the farm set and instantly feel offended by the sight of a toy pig. However, by removing the pig from the farm set could offend other parents as the pig is a farm animal and should be included in a farm set. If Jewish and Muslim parents feel offended by the toy pig then they should remove it them selves and therefore resulting in other children not being deprived of having the toy pig in their farm set. Looking at political correctness from a psychologists point of view, political correctness makes writing up a research report very difficult. The reason for this is because when writing up a research report psychologists must be very careful in the manner that they address certain groups of people so not to offend them in any way. An example of this would be when referring to a person with dyslexia. They would have to be address as ‘a person with dyslexia’ rather than ‘a dyslexic person’. This may not seem like a big deal but when your writing a research report and you are doubting yourself because you don’t know if what you are saying, or what manor you are saying it, may offend someone.

Another factor in which politics is affecting the science of psychology is the way studies are conducted now-a-days. Research studies must be approved by the ethics committee before it can be conducted as there are many things that are unethical to conduct. An example of this would be Milgram’s study on the effect of obedience on behavior. He had participants thinking they were giving a series of electric shocks to other participants which were in fact actors (read the full paper here ). A study like this would never be approved now-a-days because of ethical issues and the effect the study had on the participants. Ethics are there to keep participants safe and so dangerous or damaging studies can not be conducted. However, if we did not have ethics, research studies could become very interesting as long as they were safe.

To conclude, politics are having a negative effect on the science of psychology. However, because psychologist have to be careful in the way they address groups of people they are minimizing the risk of offending anyone which is positive. Also without ethics who knows what kind of research studies would be conducted around the world therefore having ethical guidelines set by the British Psychological Society is a good thing at the end of the day. Therefore although politics is affecting the science of psychology it is keeping participants safe.

Is it dishonest to remove outliers and/or transform data?

Outliers are the data points that stand out and are out of the ordinary compared to the rest of the sample. They can occur by chance, measurement error and sampling error. Outliers can cause many problems for researchers as they can affect both the descriptive and inferential statistics of the sample. With the descriptive statistics they can cause the mean to either increase or decrease and will increase the estimate of variability. To fix the problems outliers cause us there are three solutions. The first is to use stronger statistics such as using the median instead of the mean. The second would be to transform the data. And third is to remove them from the sample.

Removing outliers from sample’s are a solution depending on the background and basis of the research. If there was a situation where the participant matched the criteria and had simply scored an extremely different score from the other participants then removing the outlier would be wrong. The reason for this is because the participant has fitted into the criteria just like all the other participants but has simply done much better or much worse than everyone else.

Now to consider a different situation, if participants had taken part in a reaction time test and there was an outlier in the sample because one participant had got bored and just held his/her finger on the button until the end of the test, then this is when it would be appropriate to remove the outlier. The reason for this is because the data point would not be valid as it is not a score of reaction time and simply just a score of someone holding their finger on the button. Removing the outlier would give a clearer picture of the real data collected.

To sum up, removing the outliers is completely up to the researcher. In some situations it would be very dishonest to remove them; however, in other situations it would be appropriate for them to be removed.

Do you need statistics to understand your data?

Statistics is not a necessity for the understanding of data; statistics are very useful to have to get a more in-depth knowledge about the data you have collected but not a necessity. Therefore, to get the strongest possible outcome from your data, statistics is needed. Statistics can give you a firmer conclusion than what raw data can give you. Having statistics can help to conclude if your data is significant or insignificant and whether the data collected correlate with each other, for example. When using statistics it is possible to get two types of useful information at the end. The first is descriptive statistics, these are the mean, standard deviation, degrees of freedom, etc, and are used to describe and round up the data. The second type of statistics used to give us more information are the inferential statistics. These are calculations such as f-ratio, t-test and p values.

On the other hand we can have a good understanding of our data without using statistics. One way to look and summarise data without the use of statistics is to plot a scatter plot. When a scatter plot has been plotted two main factors can be concluded without any use of statistics. The first is if there is any correlation within the distribution. There can be five types of correlation, high positive, high negative, low positive, low negative and no correlation. With this information we can see the relationship between the two variables. The second factor we can see after a scatter plot had been plotted is if there are any outliers in the data. If there are outliers we can then think about going back and removing these outliers from the data to give us a stronger results. By using a scatter plot and seeing the correlation between the variables gives us useful information about our data.

To conclude, statistics are not necessarily essential for the understanding of data, however when a much more in-depth and informative result is needed statistics are vital.