December 2012 Archives

Google Analytics Tutorial

| No Comments

http://www.youtube.com/watch?v=XZDUWd_ezcI

This is a very useful video explaining how to use the new Google Analytics. As a visual learner, I really appreciated how the tutor takes us through the various pages and explained each of the percentages. As I start my internship this next summer I will definitely be watching these sort of tutorials to prepare me for all the information gathering I will be doing. Additionally, I appreciate all the information that is provided by this one program, it is almost a one-stop-shop for information. However, I would also want to cross check this information with another analytics program to ensure validity and reliability.

The Pros and Cons of Open Ended Questions

| No Comments

http://www.surveymonkey.com/s/DFDB6C8

In looking at this attached survey on bottled water I immediately noticed how the person conducting the survey did a good job avoiding leading questions by making them all open-ended. However, while open-ended questions have their advantages: respondents feeling that they have more control, gaining new information and insights, and learning the nuances of the responses. However, a major disadvantage is that the answers must then be coded, requiring additional manpower, money and time.

An additional worry that I have for this survey is the sheer quantity of open-ended questions. As I took this survey I was overwhelmed by the amount of time and energy it was taking for me to complete the survey. Several times I was tempted to quit and move on to other projects, leading to the potential for a high nonresponse rate.

Finally, any results gathered from the survey has the potential to be non generalizable depending on the variety of the responses provided. This sort of qualitative research limits the extent to which one is able to understand the extent of the phenomena. Any analysis of the data would be an interpretation of the responses and subjective to one's point of view.

The Trap of Pseudo Polls

| No Comments

Every day I come across a multitude of pseudo polls. One, which stuck out, was that of the Patch.com news sites, specifically the BarringtonPatch.com. This pseudo poll begged the question "Should uniforms for the U.S. Olympics athletes be required to be made in the U.S.?" With only two choice options -- yes or no -- and only 39 responses, the results of this poll have no statistical or scientific reliability, and are therefore not a proper indicator of the population's opinion of an issue. The unrepresentative population sampled in the survey causes the results of both polls to be invalid and unreliable. However, I believe even if the poll were to have a representative random sample, the results would still be invalid due to its limited number of response options - some people may not be fully aware of the circumstances surrounding the issues and/or other may not have any opinion.

While I believe the questions are clear, the article preceding the pseudo poll introduces bias by citing how "citizens are up in arms," and how the "reaction [was] so strong that the company pledged to manufacture uniforms for the 2014 Winter Games in the U.S." Additionally, the pseudo poll provided an option to 'View Results' prior to voting, enabling people's opinions to be swayed by the voting of others. If I could make any corrections, I would suggest that the poll conduct a random survey - meaning every member of the population has an equal chance of being chosen - so that their populations many be well represented and generalized. I have three additional suggestions for the pseudo poll: (1) conduct the poll independent of the article, (2) disable the 'View Results' option until after the poll has been closed and the results have been analyzed and finally and (3) include the additional options of 'not aware of the controversy' and 'aware of the controversy, but have no opinion.'

http://barrington.patch.com/articles/poll-make-olympic-uniform-in-us

"I have a favor to ask of you. Would you please fill out this dental survey for your awesome dentist? But I have one request . . . if you feel you cannot give him all 5 stars, please do not fill out the survey but instead email us your opinion. Constructive feedback is very good and he promises no (additional) pain at your next appointment. :-) I'm starting this with family and then will send to other patients.

This is what you need to do:

Google: Peter J Shore, DDS
Then click on
Patient Surveys for Dr. Peter J ...


Thank you so much for your time and help! We greatly appreciate this. (Be forewarned, I may pass on more surveys as I find them.)

Nancy J Shore
Shore Family CFO and Assistant Director of a whole lot of nothing :-)

Above is a survey I received from my aunt regarding my uncle/dentist. If I take anything away from this class, it is gaining an understanding the importance of ethics in research. When I received this email I followed the link to a popular website used for ranking healthcare professionals and their practices. Based on what I have learned in this class about research ethics, I knew that filling out the survey would reflect my bias due to my familial relationship with my uncle. While before I might have filled out the survey believing that it could only benefit my uncle, I now understand how my participation in the survey would only skew the results.

Additionally, I found my aunts mention of how we should not fill out the survey if we do not feel he deserves all five stars to unethical. While I understand my aunt's intention to help my uncle and build his reputation, I don't believe that she should have gone about it in that matter. Rather, his reputation should be able to stand on its own and not rely on biased input.

The Struggles Of Writing A Survey

| No Comments

For my JOUR 5501 class I had to work with a group to create a survey related to sports. Our group chose to focus on the controversy surrounding Ralph Lauren's 2012 Olympic Ceremony outfits. The survey can be found by following these two links: Part I and Part II

Successes:
Our research question asked if United States citizens had an opinion on whether the US Olympic Ceremony clothing should be made domestically. Inspiration for such a survey was the result of the recent controversy surrounding Ralph Lauren's 2012 Olympic Ceremony outfits being made in China. I feel that one of our primary successes was the way in which we devised our questions and formatted our survey. I believe that our survey's questions did a good job of collecting valuable demographic information and gauging peoples' awareness of the controversy and accumulating their opinions for future use in similar situations.

Failures:
While I have strong faith in the responses we collected, I believe that it provides a very limited point of view of the issue. As discussed in part two of this assignment, we failed to survey a representative sample of the United States population, and therefore cannot generalize the results. Much of this was due to limited time and resources, but additional problems were the result of the format in which we sent it out. Due to constraints set by SurveyMonkey, the online survey service we used to collect our results, we had to split our survey up into two parts: part one surveyed demographics, while part two asked questions related to our research question. Of the 51 people who started the survey, only 49 completed it. Ideally, all questions would be collected in one survey and our subjects would have been chosen at random, where every member of the population had en equal chance of being chosen. Without this random and representative set of subjects, we are not able to generalize our results and therefore cannot properly answer our research question.

Challenges:
In working on this project, I found the greatest challenge to be working with my group and dividing up responsibilities. To design, distribute, collect and analyze a survey with a group of people with vary different schedules and writing styles proved to be difficult. Additionally, some group members were more active in contributing to the success of the project than others, resulting in an unrepresentative approach to the study even within our group. While I back the writing and design of our final survey, I believe that the same results could have been achieved much more quickly if we had all clarified our goals and intentions before beginning the project.

Corrections:
If given the opportunity to make corrections, I would first ensure that my team members and I are all on the same page in terms of our goals and commitments for the success of the survey. Once established, I would suggest that each be responsible for surveying random subjects of various demographics in an effort to generalize the results and lower our percent error. I would also turn our two-part survey into one congruent survey to prevent people from quitting half way through the survey.

Lessons:
The primary lesson that I learned from this project is the importance of being on the same page as those who I am working with to create the survey. Additionally, I gained insight into the amount of planning and time it takes to create an effective survey that can be distributed, analyzed and generalized for a target population.

About this Archive

This page is an archive of entries from December 2012 listed from newest to oldest.

November 2012 is the previous archive.

Find recent content on the main index or look in the archives to find all content.