We're moving from DESCRIPTIVE statistics to INFERENTIAL statistics which deals with statistical inference (hypothesis testing/correlation) meaning predicting outcomes with specific levels of confidence. The text describes statistical inference as making "accurate decisions about parameters from statistics." (Kozak, 2014, p. 202); in other words, drawing conclusions about populations based on samples.

Data collection is the key, followed by data handling. Are the samples "representative" and truly random? How are "outliers" or unusual data handled? Would a pharmaceutical company throw out (ignore and not report) the 1 death in 10,000 for a new medication? NOT LIKELY. Listen to any drug ad: while happy people are enjoying life because they used this new drug, a voice in the background is listing ALL of the side effects that their research has noted - with the last one often being " . . . including death." Ethics are involved.

FLIPPING COINS

• Assume the probability of Heads is 50% (0.500) and Tails is 50% (0.500) This make the Null Hypothesis, Ho: Heads = 50% (The Null hypothesis MUST include the equals sign)
• The Alternate Hypothesis can be that the probability of Heads is Less Than (<), Greater Than (>) or simply Not Equal to 50%
• Landing on the "side" is NOT an option and would constitute invalid data and be deleted.
• Coins can NOT be badly worn down. For example, a coin with a worn-down heads side would favor Heads since the Tails side would be heavier. Using such a coin for gambling would be UNETHICAL.
• Samples MUST be randomly collected with every sample having an equal chance of being selected and demonstrating the hypothesized results (Heads or Tails).
• "Convenience" sampling is to be avoided?
• Assume that "flipping" or shaking and dropping coins onto a hard surface DOES provide adequate randomness.

1) Select ONE of the below sampling protocols and follow it AND explain why you picked that protocol and what might be biased with the 4 others.

• Use one relatively new coin (2010 or later) and flip it 30 times (sample size = 30)
• Use one old coin (from the 1940's or 1950's) and flip it 30 times
• Use 30 relatively new coins (2010 or later), Mixed values (e.g., pennies, dimes, nickels, quarters). Shake them all in your hands together and drop them onto a SHAG carpet.
• Use 30 relatively new coins (2010 or later) , ALL the same value (e.g., all pennies) Shake them all in your hands and drop them onto a hard surface.
• Use 30 old/new coins (1970's and later), Mixed values (e.g., pennies, nickels, dimes, quarters). Shake them all in your hands and drop them onto a hard surface.

2) Tell which denomination coin(s) you used (If not a U.S. coin, what U.S. coin is it closest to in size, weight) Record how many heads and how many tails you got in your flips:

_______ Heads, _______ Tails and what % these are of the 30 tosses: Heads ____% and Tails ____ % (These must add up to 100%) This is basic math.

When I receive your sampling results, I will combine them to see how close we got to the theoretical probabilities as well as point out the shortcomings of 4 of the above 5 sampling protocols. We will complete the testing of our hypothesis later in the course.

Ethics is a critical consideration in statistics. The math can analyze any data set, so it's the collection and handling of data that determine the ethics of the final statistics or predicted probabilities.

Many professional organizations publish ethical guidelines about the use of statistics and research that have particular relevance for their fields and professions. TheAmerican Statistical Association (April 2016) published Ethical guidelines for statistical practice, a report prepared by the Committee on Professional Ethics of the American Statistical Association, Alexandria, VA. Retrieved from: http://www.amstat.org/ASA/Your-Career/Ethical-Guidelines-for-Statistical-Practice.aspx

3) After reviewing this report, answer the following TWO questions related to YOUR present or anticipated field of work, AND the THIRD question on conspiracy theories (add these answers to your coin toss result post):

• Whyis it important that statistical research follow ethical guidelines?
• What aresomepotential consequences of not following ethical statistical practice (i.e., results used improperly, etc.)?
• What is a conspiracy theory you have heard about that is based on biased data collection or handling that you can provide some statistical evidence to refute?