MATH 533 Week 1 Discussion: Study and Sample Design – Download To Get An A+

Document Content and Description Below

MATH 533 Week 1 Discussion: Study and Sample Design – Download To Get An A+ Week 1 Statistics Discussion Keller Graduate School of Management; Professor Aaron Marmorstein Weekly Discussion, 3-6-... 2019. “A company makes men's button-up dress shirts and is considering changes to its design. How might the company go about finding a good sample of individuals? These decisions would include method of sampling, location of sampling, population to consider, and number in sample, in addition to any other considerations. Here's the basic idea: I want to perform cluster sampling at three of my 10 stores. I randomly select three stores (or choose three in different socio-economic locations) and ask an employee in each store to watch for clients shopping for shirts. The clients can then be asked specific questions on what they're looking for and what they like or dislike about specific designs. If possible, you can ask about age or career or other attributes that you think are important. You might even consider offering the client’s $5 gift certificates for their time. In each store, once 30 client’s preferences have been gathered, we should be in a position to start analyzing the results. Do you think this would work? What would you suggest differently?” That type of sampling is very productive in results, because the whole population is available, where a stream of representatives are on a face to face instant response to you, as the researcher. It's sort of like the random selection hat joke where the guy say's he puts all of his monthly bills in a hat' and the one that he selects one from. That's the one that will get paid this month. Professor and Class: I would like to agree with the idea of random sampling, but--THERE IS ALWAYS A BUT!!! is Simple Random Sampling the better of the two versions of random sampling (i.e., Simple and Stratified) I thought the article you posted highlighted some of the reason, to consider the pros and cons of both. I reviewed an article commenting on the advantages and disadvantages of Simple vs. Stratified. I guess we would have to take into consideration the basis and requirements of the sampling first, like whether there is a quota requirement. Since it would involve participants that you can observe on a more personal face to face level, there are qualitative elements involved, right? Sometimes face to face can mean you might want to carefully select the types of questions you're going to be asking. If any are of a person nature, there is the possibility of making your respondent feel embarrassed. Plus, what are the psycho graphics. what type of person or people or we targeting--couples, parents, seniors, students, swingers, spiritual, yuppies, guppies, hippies--well you get the picture and demographics of the target respondents? On the demographic end of things, what lifestyle and social levels will be the targets, low income, affluent, and what area of living standards. All that comes into play. I suppose the great advantages of both are that you can control certain aspects in the questioning by being able to detect when people are being honest and or ethical in their responses...or just plain "silly". You can also determine bias and emotions in expressions and body language. Additionally, you get to select your targets right on the spot, and you receive instant responses, so you get the contact points right away--the entire quota. Thing is, as you suggested Professor, Marmorstein, people sometimes need incentive to want to give you even a moment of their personal time, but you have to beware of ambiguous quick responses just to get the reward. Are the respondents being truthful or giving you their honest opinion, or just saying what they think you want to hear. If our sampling is qualitative, we want truth and real personal opinions...even if the question is of a sensitive nature to the respondent--or to yourself even. Plus, as the professor's article suggested, you don't want to trap the target into a controlled response, so we've got to remember to leave the questions open-ended and allow room for honest emotional response. [Show More]

Last updated: 1 week ago

Preview 1 out of 3 pages