Introduction to Statistical Reasoning

"Statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write." H. G. Wells, (1866 - 1946)

Evaluate the study "Residual neurocognitive features of long-term ecstasy users with minimal exposure to other drugs" in terms of the seven steps outlined in chapter 6 of text. Note: For step 2, you must critique the study in terms of the seven critical components.(3-5 pages - 80% of grade).

Critique the related news article's (titled "Ecstasy does not wreck the mind, study claims") presentation and interpretation of the study findings, using the review criteria outlined at Health News Review (1-2 page - 20% of grade). Note: If you think a review criteria is "not applicable", please state the reason why.

Project 02 (due April 16 - worth 8% of overall grade)

Design, execute, and write up a sample survey on a topic of interest to you. Design your questionnaire to have 10 questions. Include at least 30 people in your survey. Survey Monkey is a useful tool for designing your survey.

• Include questions that result in both quantitative data(Example: age, income, blood pressure, temperature) and categorical data, both nominal (Example: Gender(Female/Male), Race(White/Black/Others)) and ordinal (Example: strongly agree, agree, neutral, disagree, strongly disagre).
• Include at least two questions that result in quantitative variables that you feel may be related. (Example: time studying and homework score) Note: Do not use weight and height
• Include at least two questions that result in categorical variables that can be displayed in a 2x2 table (Example: Male/Female - Yes/No).
• Include at least two questions that result in quantitative variables for making mean comparisons between groups (Example: Compare mean length of time studying per week for Males and Females)

Writing up your Results:

1. Purpose of the survey.
2. Summary of findings and conclusions.
3. Description of how the data was collected.
4. Detailed description of your findings plus summary statistics (mean, median) and graphical presentations (histograms/bar graphs) for the 10 questions of interest. Discuss the core quantitative variable questions in terms of central tendency, making comparisons where appropriate. For the categorical variable questions resulting in 2x2 tables, discuss the differences in the response variable (in terms of proportions) for the categories of the explanatory variable.
5. Limitations of survey and data collected (for example: convenience sample, nonresponse, problems with wording of questions, small sample size, missing values, outliers, non-normality of data)

Note: The write-up should be 4 to 5 pages not including graphical presentations. Note: Click here to download tips on completing surveys. Click here to add your survey link to a Google doc so other students can take your survey.

Project 03 (due April 30- worth 6% of overall grade)

Using the data you collected in project 2, complete three separate statistical analysis as follows:

1. Complete a linear regression analysis looking at the relationship between two sets of quantitative data. Clearly state what the slope of the regression line means. Create a graphical presentation of the data with the linear regression line included. Complete a hypothesis test to test whether the true population slope is statistically significantly different from zero. Note: Here is a link to how to get the slope and standard error of the slope for the regression line in Excel.
2. Complete a two-sample hypothesis test for the comparison of means between two groups (Example: Male and Female). Construct a 95% confidence interval for the true population mean difference, clearly explaining what the interval means.
3. Complete a hypothesis test for independence in a 2x2 table (Ex: Male/Female - Yes/No).

For each of your hypothesis tests, use a 5% alpha-level of significance. Clearly follow the steps for making a statistical decision explaining what the p-value means. State your conclusion in terms of the original question of interest. What type of error (type I or Type II) might you have made in your conclusion? Discuss the consequences of making each type of error. Discuss the validity of the statistical analysis you are using in terms of quality of your sample data and sample size.