Discussion Question #3: Think about and discuss skewed distribution? what causes
ID: 3071790 • Letter: D
Question
Discussion Question #3: Think about and discuss skewed distribution? what causes a skew in statistical terms? And how does one deal with skewed data when conducting research? Are there specific types of research questions and types of data where one would expect the data to be skewed? For example, if we have a skewed distribution in a variable, would we still want to include it in a statistical analysis? Why or why not? Which variables are more likely to be skewed than others? Discussion Question #4: Describe how the wording of survey questions may skew the data. For example, would asking about interest in gender studies courses vs. interest in women's studies courses influence the responses? Are there any conditions in which it would it be acceptable to allow skewed variables into a research study? If so, describe these conditions.Explanation / Answer
Question 3
Answer:
We know the term skewness is used to measure the symmetry of the probability distribution of the random variable about its mean. In positively skewed data, the mean is usually greater than the median and in negatively skewed data; the mean is usually less than the median. In skewed data, we found that tail of the distribution (right side end or left side end) is longer.
Sometimes data contains extremely large observations or extremely small observations or outliers as compared to other observations in the data set. These outliers cause a skew in statistical terms.
It is important to handle skewed data very carefully during the research study because it may produce biased results. We can use the median as the measure of the central tendency if the data nature is extremely skewed. Also, we can eliminate some outliers from the data and then we can use this data for further analysis.
There are specific types of research questions and types of data where one would expect the data to be skewed. For some specific survey questions, respondents give the extreme point response. Sometimes we get overestimated responses or underestimated responses because respondents do not wish to share real situation.
We can include skewed data in the analysis because it is important to check the facts behind it.
The variables for which mean, median, and mode are not the same, these variables are likely to be skewed than others. For example, the variable income of the person has the skewed nature.
Question 4
Answer:
If we do not use proper wording in the survey questions, then there is a possibility that the respondents will hide some information or we will get overestimated responses or underestimated responses from the respondents. Respondents may hide real situations and s/he does not wish to expose all information. There would be a gender gap while receiving responses from survey questions. Effectiveness and accuracy of the survey questions are based on the proper wording of the questions. Sometimes people answer the question which is more socially acceptable. These types of responses produce bias which is called as socially desirable responding or SDR bias. One can ask a survey question in a neutral way of avoiding biases in the study.
In some conditions, we can allow skewed variables in a research study if there is no any other option available for reducing the skewness of the data. Actually, some variables have skewed naturally and we can’t remove complete skewed nature of the data but we can reduce the effect of skewness by using some proper methods.