AP Stat Soup Survey Project: Writing Surveys

Here's the information presented on 12/7 regarding survey writing basics.

Survey Basics

• Decide what information you want to gather from the survey.
• Ask only those questions that will provide the information you need.
• Keep the survey as short as possible
• Use multiple-choice questions whenever possible.
• Avoid leading and biased questions.
• Use the same rating scale throughout your survey.
• Test the survey before you roll it out.

Source: http://www.ehow.com/how_16596_write-survey-questionnaire.html

Survey Writing Specifics
Directions to Respondents
   1. Include a brief explanation of the purpose of the questionnaire.
   2. Include clear explanation of how to complete the questionnaire.

Content of Questions
   1. Will the respondent be able to answer your question, i.e., do they know the answer?
   2. Will respondents want to answer the question, i.e., is it too private or silly?

Wording of Questions
   1. Will the respondent understand the wording?
   2. Are any words so strong that they might influence the respondent's answer?
   3. Are you asking more than one question at a time by using "and" in your question?
   4. Do multiple choice question choices mutually exclusive and encompass the total range of answers?

Order of Questions
   1. Be careful not to include so many questions that potential respondents are dissuaded from responding.
   2. Attempt to get recruit respondents' motivation to complete the questionnaire. Start with fact-based questions and then go on to opinion-based questions.
   3. Attempt to get respondents' commentary in addition to their ratings.
   4. Put a date on the form so you can keep track of all future versions.

Source: http://www.managementhelp.org/evaluatn/questnrs.htm

Scales in Surveys

• When using scales, create a balance between positive and negative options.
• An even number of choices forces the respondent to decide whether they lean to the positive or negative.
• An odd number of choices allows the respondent to select a neutral response.

Example: How does the flavor of this soup compare to the brand you usually eat? The flavor is

1) Much Worse  2) Worse  3) Slightly Worse  4) Slightly Better  5) Better  6) Much Better
(This example shows the use of a six point scale that forces the respondent to make a decision as to whether or not the soup tastes better or worse.)

1) Much Worse  2) Worse  3) About the Same  4) Better  5) Much Better
(This example shows the use of a five point scale that allows the respondent to be neutral as to the taste.)

Additional information and resources can be found at:
http://4h.uwex.edu/evaluation/documents/Wordingforratingscales.pdf or
http://www.gifted.uconn.edu/siegle/research/Instrument%20Reliability%20and%20Validity/Likert.html

AP Stat: ESP Investigation Guidelines

For the Chapter 11 ESP Investigation, be sure you do the following:

1. Describe how you are conducting your simulation
2. Describe what your response variable is
3. Report the results for each trial
4. Summarize your results
5. State a conclusion that is backed by your analysis

AP Stat: US Census Population Totals

Here's the US Population data that we looked at in class. You are to re-express the data so that it yields an appropriate linear model. Please be sure to check your residuals.

2010 310,232,863
2000 281,421,906
1990 248,709,873
1980 226,542,199
1970 203,302,031
1960 179,323,175
1950 151,325,798
1940 132,164,569
1930 123,202,624
1920 106,021,537
1910 92,228,496
1900 76,212,168
1890 62,979,766
1880 50,189,209
1870 38,588,371
1860 31,443,321
1850 23,191,876
1840 17,069,453
1830 12,866,020
1820 9,638,453
1810 7,239,881
1800 5,308,483
1790 3,929,214

AP Stat: Smoking Investigation Feedback

You are to comment on at least two reports in the discussion section of the AP Stat wiki.

Here are a few things to look for when giving feedback on the Smoking Investigation report:

1. Were the W’s addressed in the paper?
2. How did the person demonstrate that a linear model is appropriate?
3. Are graphs labeled and understandable?
4. Was the model described in context and written using proper notation?
5. Do any predictions made using the model look reasonable?

AP Stat: Create a Regression Equation from Summary Statistics Data

Here are the summary statistics and correlations for the Arm Span and Pass the Buck activities:

	ARM SPAN	vs	HEIGHT		TIME	vs	STUDENTS
MEAN	69.668		69.082		5.305		10.879
Standard Deviation	4.011		3.214		2.756		5.343
CORRELATION		.885				.950

Create regression equations from this data and write the meaning of the slope in context of the problem.

AP Stat: Survey Project Data

If you have trouble reading the Excel file containing the survey data, I have posted the data as a text file. This data can be cut and pasted into the online graphing applets that are available through the AP Stat links page.

AP STAT: Model Parameters

I thought it was important as your are going through Chapter 6 and moving forward to keep clear the differences between distributions and statistical models.

The difference between a distribution and a model:

Distribution              Model

real data                     theoretical values

observed                    imagined

histogram                   mathematical curve

statistics                     parameters

center y                      center μ

spread s                     spread σ

AP Stat: Three problems from class.

Here are the three problems from the presentation that you are to complete by Friday.

3) How many states would you have to visit to have a z-score of 1.5? (Use the class mean and standard deviation from your class survey to answer this question.)

4)  Data on the weights of cocker spaniels has a mean weight of 21.2 pounds. A cocker spaniel weighing 18.3 pounds has a z-score of -1.1. What is the standard deviation for this data?

5) Data for one fifth grade class' long jumps has a standard deviation of 1.6 feet. One fifth grade girl who jumped 7.8 feet has a z-score of 1.3. What is the mean distance jumped for fifth graders in this class?

AP Stat: Comparing Data Distributions

When Comparing Centers
Use the same measure
    - means vs means
    - medians vs medians
Describe differences
    - which is bigger, smaller
    - how much difference is there

When Comparing Spreads
Use the same measure
    - std dev vs std dev
    - IQR vs IQR
Describe differences
    - which is broader or narrower
    - how much difference is there

When Comparing Shapes
Note symmetry differences, outliers, gaps, and unusual features

AP Stat: Online Tutorials and Exam Preparation

I came across Stat Trek and thought it could be a valuable resource as you progress through this year's course.

If you go to the tutorial, we have covered the first session under Exploring Data and are in the middle of the second session. This is just another presentation of the same material, so you may find it helpful to have a different presentation if you don't understand something from class.

AP Stat: Constructing an Ogive

We were a little rushed at  the end on the actual Ogive construction, so I put together a series of images to illustrate how to finish off the Ogive.

This first illustration shows the Cummulative Percentage Histogram, which is what we had in class. I am using the M&M data posted on the blog and looking at how many M&M's each color has in a bag.

Next, put a dot at the end value of each bin. You can place the dot at the position of the maximum value in each bin to be more precise.

Finally, erase the boxes and connect each dot with a line.

You now have a rough curve of the data distribution. This graph is your Ogive.

Please note that you can construct Ogive graphs using cummulative counts rather than cummulative percentages. It just depends on what you want to say about your distribution.

AP Stat: Podcast for Chapter 4 Posted

The Chapter 4 podcast is now available. You should listen to the podcast to be sure you understand all the important elements of the chapter.

AP Stat: Study Habits Article

I saw this article in the Denver Post a couple of days ago. Besides citing research studies (yes, another example of a statistics-based article), the article provides information on cognitive learning research and ways to maximize the effectiveness of your studying. I think it has some good suggestions to help you maximize your learning.

Forget what you know about study habits, Denver Post, 9/9/2010.

AP Stat: Summary of Writing Statistical Content

I thought you all might like to have the Wordle that I created from the comments on the statistics-based newspaper article. It really is a good summary of what you should be doing in any statistical writing.

AP Stat: Calculators in Class

Now that we are starting to look at graphs for quantitative data, you need to be sure to have your graphing calculator in class every day. We will also be working through more problems from the book, so you should bring your textbook to class as well.

AP Stat: Chapter 3 Podcast now available

The Chapter 3 podcast is now available. You can use this as a review vehicle to be sure you understand all of the important concepts in Chapter 3.

AP Stat: Writing a Good Question Response

Here's what was presented in class today:

When writing responses to questions

• Re-state the question in your own words
• State the facts you know about the problem
• Describe what you will do for your analysis
• Perform your analysis and report your results (graphs, tables, calculations)
• Write your conclusion about the question

AP Stat: What to Include When Writing a Statistics-Based Newspaper Article

The following are the thoughts generated during class on what to include when writing a statistics-based newspaper article.

Period 4

• Beginning - address the what and who
• State percentages within article in context
• Break down data, explain how data is described in graphs
• Explain what features are shown in graphs and tables
• End with a conclusion of what point you are making
• Include absolute numbers to give another perspective of size

Period 5

• Use numbers and percentages that can be compared
• Provide historical reference when possible
• Come to a direct conclusion
• Use graphs and charts to help compare
• Title should be relevant to topic
• Provide analysis as to "why"
• Explain graphs
• Write without bias or opinion

Period 6

• Explain how data was obtained
• Provide graphical representation (table or graph)
• Describe what happens with data without bias
• Make title tie explicitly to article conclusion
• Tell when data collected
• Explain data thoroughly (tell what numbers apply to what)
• Say why data matters
• Provide historical context (if available)
• Compare similar sized samples
• Organize presentation in order of time (don't jump around)

AP Stat: M&M data for classes

Here are the data for each class. You are to graph/visually display your data and create a contingency table to determine whether or not the color of an M&M is independent of its type (plain or peanut). Come to Monday's class prepared to discuss your conclusion.

 Period 4

		Green	Brown	Blue	Orange	Red	Yellow	Total
PLAIN		161	84	206	181	82	115	934
PEANUT		72	31	57	52	30	51	293

 Period 5

		Green	Brown	Blue	Orange	Red	Yellow	Total
PLAIN		211	115	207	211	115	131	990
PEANUT		65	41	84	62	52	72	376

 Period 6

		Green	Brown	Blue	Orange	Red	Yellow	Total
PLAIN		193	173	240	237	137	176	1366
PEANUT		83	42	93	91	43	91	443

AP Stat: Chapter 2 review

You can listen to the Chapter 2 podcast to help you review key concepts for the quiz tomorrow.

Preparing for First Day

For the first day of class you should come prepared to discuss any questions regarding the summer reading and homework assignment. Your homework will be collected on Monday. On Tuesday we'll have a quiz. So get yourself ready.

I have a podcast for Chapter 2 that highlights important points that you should understand. If you don't, you should go back through Chapter 2 and come to class with any questions.

Have a great weekend; see you in class on Monday.

Setting things up

I am preparing online resources for the fall semester. There will be little or no activity on this blog until late August. But, then again, you never know.