Answer the following discussion questions:
1. Comment on the possibilities of "lying" with or manipulating statistics by using very small, very large, or non-random samples.
2. Do you know of any studies or surveys where you suspect that these tactics have been used? Find an example and share it with the class.
3. If you were opening a new business, what kinds of surveys or studies would you want to conduct before investing your time and money? How would you conduct these studies?
Make a substantive reply to the comments of at least two of your classmates. Please respond to the merits of a student's comments.
1 - Statistics seem a field where even accidentally through a poorly designed study, draw inappropriate conclusions based on the data gathered. Other times hiding the methodology of the study is tantamount to lying. If the study creators knowingly compile a very small sample or ask only certain people or people in only a certain place and make sweeping claims about the population, then yes that is most certainly a dishonest manipulation of the data.
I don't know of any such surveys personally, but it seems evident especially in an election season all the voluntary call in or text in polls a tv program posits are inherently skewed. Those that watch CNN aren't commonly going to watch Fox News and vice versa so not only is it a voluntary study, it is narrowed to only those watching said program and often will hold considerable bias.
I think for a new business survey I would be looking at the population makeup. Does the area contain the age demographic that would most often buy my product or service? How about median income? I wouldn't open a Rolls Royce dealership in a middle class town etc...
Are there other competing businesses in close vicinity? I'm no entrepreneur so I'm not sure what kinds of personal questions would be useful. How someone votes or feels about religion etc are probably not that useful if you want to open something general like a coffee shop or something. Most of the data I would think to consider is demographic information the city/county/state can provide.
2 - It is very possible to sway the results of a survey by using a small non-varied group of participants. If you as a researcher are going to theorize the outcome of your study would effect a large population, yet your sample only looks at female college students between the ages of 20-25, then you cannot tangibly apply your results to those outside of that age range or demographic or gender.
For example, in the study "Focusing on Self or Others has Different Consequences for Psychological Well-Being: A Longitudinal Study of the Effects of Distinct Interpersonal Goals." The researchers hypothesized that having self-image goals were predictive of increases in negative feelings (depression, stress, anxiety), and compassionate goals were predictive of decreases in those same negative feelings.
To determine their results, they utilized an online survey of questions with a scaled rating (1-5), to be completed every two weeks for 3 months. The researchers found that compassionate goals were not associated with depression and stress, but were correlated with feelings of being connected, closeness, and positive effects. Whereas, self-image goals were associated more commonly with depression, anxiety, and stress. However, the limited variety of respondents is a hindrance to the validity of the findings. Only 161 university students were sampled; 95% were female; the average age was 21; and, 98% were single. The lack of diversity among participants does not provide a varied scope of results among the average population. The researchers did account accounted for possible association between interpersonal goals and residual score of depression, anxiety, and stress. To further prove their hypothesis, a larger experiment group with much greater diversity is needed.
If I was opening a business, as a personal organizer, I would want surveys conducted on the following. 1. If people felt they would benefit from having someone help them organize an area of their house or life. 2. How much people would be willing to spend on organizing services. 3. If there was a time of year that they would be more likely to want organizing help ("spring" cleaning for example). To conduct these surveys I would conduct an online survey to rate the above listed criteria. The population would be gathered from shoppers of a national grocery chain store and a local grocery store chain. As my business would be locally based I would like to compare data between national results and localized results. Using participants from grocery stores nationally and locally would allow for a diverse group of respondents, from a variety of socio-economic, racial, age, and gender demographics. The online survey would consist of questions with qualitative data (yes/no questions), and quantitative data (amount willing to spend on organizing services).
References - Duarte, J., & Pinto-Gouveia, J. (2015). Focusing on Self or Others has Different Consequences for Psychological Well-Being: A Longitudinal Study of the Effects of Distinct Interpersonal Goals. Journal of Social and Clinical Psychology, 809-825.