Sunday, December 11, 2011
"Need to prove something you already believe? Statistics are easy: All you need are two graphs and a leading question"
Thanks to Tommy Turner for sharing this
Saturday, August 13, 2011
"In the end, Boyer found just two hospitals in Tucson prepared to participate. Between May 2004 and October 2005, she and her colleagues treated 15 children as part of the trial, mostly under the age of six. Even with the small sample size, the results were striking: symptoms of the scorpion’s neurotoxin resolved within four hours for all eight children receiving Anascorp, whereas only one of seven subjects receiving placebo recovered so quickly. The rest received heavy doses of sedatives and had to stay for longer in intensive care units before it was deemed safe enough to discharge them from the hospitals."
read more at
Sunday, June 5, 2011
At the same time, make sure your survey questions are relevant to the participants so as to increase response rate.
2. If you use an online survey, send it out (roughly 3 weeks)early to a large sample and be prepared for low response rate (15%-30% is not a surprise). Show people the benefits of completing your survey; you might want to include some incentives (candies, pizzas etc). Do not hesitate to send our reminders.
3. Set up a lot of numeric variables and equations so it will be much easier to find correlations in the end. Keep finding comparable data from other resources to do hypothesis testings in your final report. Give yourself a few days to work on computation and writing the report. Be clear and organized by using graphs and data supported by your own analysis.
4. Have trouble understanding materials in class, talk to your professor as soon as you can. Chulk Stull would love to help you.
Goodluck! It's a fun class if you love what you do so make a good choice ;)
Saturday, June 4, 2011
2. MAKE SURE you get a large sample, because small samples can give you data that isn't necessarily representative of the population.
3. When choosing a method of obtaining data, whether survey, direct observation, etc. be overcritical of your method and try to eliminate any bias.
4. Get as much data as possible that is related to your subject, you may find new correlations or interesting trends that will make the report much more informative.
Friday, June 3, 2011
2. Never stop in the middle of the analysis. Do a little bit every week. If you stop, it’s kind of hard to pick up the ideas why you start this and it’s kind of hard to re-remember the reason you like the topic you choose.
3. If you do the survey with the e-mail, please send much number than you expected. The percentage of the people who will answer the survey though email is much smaller than you expected.
4. If you find something goes wrong during the research, talk to Dr.Stull as soon as possible. The suggestions will be really useful and it will save a lot of time doing nothing.
5. Besides, Good Luck!
The research project is a fun experience. However, you will have moments when you love doing it so much and other times when you would like to throw it out your window. It is important to carefully choose your partner and come up with a topic that both of you like. When designing the survey make questions that are numerical and that will give you enough data to compute, do hypothesis testing and draw significant conclusions. Also, when doing an online survey, be careful with the websites as you may think they are for free, but after collecting some few responses they will ask you for money in order to access the answers. Finally, but not last, take a lot of time in order to do this. Believe me, it takes longer than you expect and you do not want to be writing the final report few hours before the deadline.
If you're reading this advice page then you are already well on your way to being successful in this class. Believe it or not this term will fly right by you so here is some advice I have for keeping up with the pace.
1. Keep it simple!...the simpler your project idea, the easier it will be to interpret results and understand your findings.
2. Research something you are really interested in, the more interested you are in your project the more thought you'll put into collecting your data and analyzing your results.
3.This project is extremely time consuming so the best thing to do when collecting data is to keep track of information as you go along, especially if you're working by yourself.
This project taught me a lot about patience and the importance of direct language. They way you word a question has a tremendous impact on the types of responses you get and the work you have to put into analyzing those responses later on. Also, try to think of creative ways to get your surveys out to people. Try to weigh the pros and cons of each method of collecting data. For example, my partner and sent our surveys through random emailing and collected our data online,this saved us some time, but we were not able to get 100 responses so our final grade suffered.If I could have I would have asked more questions about how to interpret the numbers because correct calculations aren't enough if you don't know what they mean. Looking back I would have followed all of the advice I am giving you now and I would have received a better grade as a result.
I think the most important thing for this project is for students to create their survey and hand it out as soon as possible. This leaves more time for the student(s) doing the project to toy around with the data and try to decipher as much useful information as they can out of their data. In particular, regression is an important device that students need practice at in order to refine their skills regarding analyzing regressions. The more time students leave between the time they finish the survey and the actual due date for the project, the less stressful (and more enjoyable) the project will be.
Our data yielded interesting results that are very revealing about Kalamazoo College. Using regression, I was able to determine whether there were positive correlations between satisfaction level and other factors, and the importance of certain factors that we tested regarding a student's preference in choosing Kalamazoo College.
Our findings clearly indicated to us that students who chose Kalamazoo College for academics or for sports reasons were, on average, highly satisfied with Kalamazoo College whereas those students who chose Kalamazoo College for financial aid reasons were, on average, much less satisfied. Other interesting finds that we discovered was that a student's GPA had no correlation with the student's overall satisfaction level with the school, indicating that GPA was not a determining factor in overall satisfaction with the school. Also, we were able to determine that student's who preferred to attend a school closer to home tended to be far less satisfied with Kalamazoo College than those who had no preference in whether or not they attended a school closer to home, showing us that distance to home is an important factor to some students, but has no effect on others.
Thursday, June 2, 2011
The most important advice I can give is to make your survey as finely tuned as possible. It may take several drafts and test runs but it’s worth the effort in order to have a survey that asks the right questions and is clear to those taking it. This will make analyzing the data much easier for you later on and will avoid unusable responses. When you write your survey know exactly what you are looking for so you don’t forget to collect a vital piece of information. It’s also a good idea to start early and send out your survey with plenty of time for people to respond and even more time for you to organize and analyze it.
Secondly, please make sure your survey is very clear, concise, and well designed. This is really important because if you don’t have a well-designed survey your sample is not going to take your survey seriously and they will put unreasonable answers. Also, make sure you put thought and time into your survey questions and make sure they are related to your topic so you can find useful findings. If you do this, it will save you time and it will make your statistical analysis easier. In addition to the survey, please make sure you distribute them as soon as possible and start collecting data right away, since putting data in excel takes a lot of time.
Thirdly, please make sure you do every single assignment Chuck gives you. Each single assignment is part of the big project so put time, thought and commitment. Also please make sure you find several articles related to your project so you can have interesting comparisons to your findings. Finally, do not procrastinate and think positive all the time even if the project is not coming as you want. And remember this project is designed to apply what have you learn in class. Enjoy.
Wednesday, June 1, 2011
A study conducted by students in the Business Statistics class has revealed interesting findings about the travelling habits of K College students. On average, K students gather in a group of four to get off campus 8 times every two weeks. During that period of time, each student spends roughly $13.3 on transportation (about 13% of their pocket money).
Why?Among the top reasons why students get off campus, eating out (5 times a month) is by far the most popular, followed by parties (weekly), shopping (3 times a month) and volunteer work (twice a month). Nearly 92% of the survey respondents said they have gone out to eat at least once during the last two weeks. While students appear to party more often than doing shopping, the percentage of people who go shopping is actually higher with 78% versus 71%. Bars and clubs do not seem to be a good idea for parties since only one third of the respondents have visited such place in the last two week. This happens to be equal with the student participation rate in volunteer activities.
Where? For many K students, Crow’s Nest and Burdicks have been familiar names beside Panera, Qdoba or Saffron, etc when it comes to dining out. About 27% of the respondents claim Crow’s Nest to be their favorite restaurant and 13.6% chose Burdicks. There is less variation for shopping locations with Meijer (41%), Crossroad Mall and Target being the most visited addresses. Finally, if you want to chill out at a bar or club, visit Metro, Beer Exchange and Waldo’s since you are more likely to come across a few friends.
When? Friday and Saturday after 4PM are not surprisingly the time students get off campus most often. Generally speaking, more than half of the students leave campus on Friday (50.61%) and Saturday (55.43%) around midnight. It is noticeable that students get off campus on Monday from 4 to 7PM as much as they do at the same time on Saturday (34.94%).
Shuttle service & Bike renting
The study looks at how K students support a shuttle service to the airport, the train station and common places off campus in relation to their travelling habits. While nearly 43% of the students drive their own cars to get off campus and 62% ask friends for a ride or borrow a friend’s car, more than 77% of the population want a school bus service to common places off campus. Only 11% use bus, which suggests that the bus system at K might not be highly accessible to the students. They also expressed their willingness to pay $1.58 per round-trip if such a shuttle were provided. It is interesting that though 35% of the students took a plane to travel outside Kalamazoo over Winter quarter and Spring break, only 32% would like to use a school airport shuttle. Bus service to train station receives much more support at 55%.
About 66% of you might have known about the “Borrow a bike” program at DeWater yet only 12% have ever used the service. Among 34% of those who do not know about the program, a big proportion of 76% indicated that they would consider renting a bike. This once again suggests that students need more accessible means of transportation to get around and become better involved in the K community.
Given such information about the students’ travelling habits, the school government could consider implementing an efficient program that would meet the student need and encourage them to enjoy off-campus life.
Carefully design your survey and make sure your questions are both clear and concise so that more people will actually answer them. Keep in mind that people want to put in as little effort as possible when taking your survey, so questions requiring textual answers may be skipped by many respondents. Also, try to think of specific variables you would like to compare and design your survey to ensure that you will obtain the variables you need from the questions that you ask in your survey. Sorting through data and arranging it to be suitable to perform statistical tests can take longer than expected, so make sure you have ample time to do so.
Chuck also gives different assignments pertaining to the research project throughout the quarter. Putting thought and effort into those assignments will help you complete your final paper (when you complete those assignments you basically finish a section of the paper).
This is a large project and I think it is very important to find a good partner. Make sure both you and your partner clearly understand concepts learned in class (confidence intervals, hypothesis testing, regression, etc.). If you have any questions, do not hesitate to ask Chuck. It is important that you and your partner both clearly understand how to do the assignment so that you can divide the work with your partner efficiently.
Foreign language study is a prominent part of the Kalamazoo College curriculum. All students are required to take 3 units in a second language (through the Intermediate Level 103 or 201, depending on the language). In addition, 89% of students surveyed indicated that they have studied abroad or plan to study abroad and most study abroad programs require proficiency in a foreign language. Due to the importance of foreign language education at Kalamazoo College, a recent survey of 79 students at Kalamazoo College was conducted to determine student satisfaction with the college’s foreign language departments and possibly identify any improvements that can be made within the departments.
Overall, students rated that they are satisfied with the current language requirement (average desired unit requirement=2.98) and the language departments as a whole (rated 3.58 on a scale from 1=Very Dissatisfied to 5=Highly Satisfied).
Although students appear generally satisfied, the study also found support for expanding the language departments. On average, most students indicated that they had a high desire to learn languages not currently offered at Kalamazoo College (rated 3.81 on a scale from 1=No Desire to 5= Very High Desire). The study also supported that students have around the same level of desire for learning languages not currently offered at Kalamazoo College as they do for improving in languages at Kalamazoo College. Students listed 19 different languages that they would like to learn that are currently not offered, and the top 3 most frequently listed were Italian, Arabic and Russian. On average, students who were willing to pay for private language lessons or lessons in languages currently not offered by the college said they would pay approximately $115 per quarter. Therefore, it could be beneficial to conduct further studies on adding more components to the language departments.
Alternatively, rather than expanding the language departments to include more languages, more emphasis could be placed on certain aspects within the language courses already offered. Students rated that they have achieved between some proficiency and high proficiency in the languages they have learned at Kalamazoo College. However, the study found that out of 4 different areas (speaking, reading, writing, and understanding speech) students felt the least proficient in writing in foreign languages learned both before Kalamazoo College and at Kalamazoo College. These results could indicate a need for greater focus on developing writing skills within foreign language courses or the development of separate writing-intensive foreign language courses.