Question

It is a descriptive stat discussion question

Write at least three reasonable ways to deal with the below outlier at length. (number them first and explain)

In response to a question about how many "close friends" she has, one participant writes “32”

this is a discussion question about descriptive stats. write at least two reasonable ways to deal with the below outlier

In response to a question about how many “close friends” she has, one participant writes “32.”

number the resonable ways and explain at length

Answer 1

Write at least three reasonable ways to deal with the below outlier at length. (number them first and explain)

The reasonable ways to deal with the below outlier are,

1. Correct the data file if the data is incorrect
2. Gauge the understanding of the participant and exclude the outlying response
3. Use median or other statistics of the value '32' is a genuine and honest estimate from the participant
4. Generate results by including and excluding outliers and discuss differences

Correct the data file if the data is incorrect: If it just incorrect data or there was a manual error or mistake during data entry - correct the file and proceed with data analysis

Gauge the understanding of the participant and exclude the outlying response: If the participant did not understand the definition of 'close' friends, this response can be excluded (not rejected), and the reason for exclusion be explicitly mentioned.

Use median or other statistics if the value '32' is a genuine and honest estimate from the participant: Median and other such statistical methods handle outliers efficiently and do not disrupt results because of the outlier. Therefore, using this method should help using outliers efficiently.

Generate results by including and excluding outliers and discuss differences: If the results are the same when the outlier is included or excluded, then ignoring the outlier does not change the result. It can be excluded. Whereas if including the outlier causes a huge change in result, then it should be reported and the reason for differences should be analyzed.