Unit 9: Quantifying Statistical Effects |
Unit 9: Assignment #1 (due before 11:59 pm Central on MON JUL 13):
- In this Unit, you’re going to learn how to quantify statistical effects, starting with how to quantify relative positions within a distribution.
- First, remind yourself what percentile scores are by reading an excerpt from Roell’s (2019) article, “How to Understand Score Percentiles.”
- Second, cement your remembrance about percentiles by reading an excerpt from Wikipedia’s (2020) definition of “Percentile.”
- Third, learn how to calculate percentile scores using your chosen data management system by completing Andrews’ (2020) tutorial “How To Calculate Percentiles in Microsoft Excel, Google Sheets, or Apple Numbers.”
- To work with the “Sleep (in hours)” data, which you’ll need to work with for completing the tutorial, you should open a new spreadsheet and either
- type in the 18 values provided in Column A of Figure 1 of the tutorial
- OR
- copy those 18 values from YourLastName_PSY-210_Unit03_FrequencyDistribution_Andrews spreadsheet, which you created in Unit 3.
- To work with the “Sleep (in hours)” data, which you’ll need to work with for completing the tutorial, you should open a new spreadsheet and either
- After completing the tutorial to learn how to calculate percentiles:
- First, save the spreadsheet in which you calculated percentiles during the tutorial with the filename YourLastName_PSY-210_Unit09_Sleep_Percentiles
- Second, take a screenshot of the percentiles you calculated during the tutorial.
- Name the screenshot YourLastName_PSY-210_Unit09_Percentiles_Tutorial_Screenshot.xxx (where xxx is the filetype, for example, .jpg, .png, .jpeg and the like).
- Your screenshot should include only the part of your spreadsheet that contains data, NOT your entire screen.
- Now, let’s get some practice calculating percentiles on other data sets.
- First, identity which four datasets you will be working with during this assignment by identifying the first letter of your LAST Name.
- If the first letter of your LAST name is A through M, choose the data sets that are marked
**A-M Data Sets**. - If the first letter of your LAST name is N through Z, choose the data sets that are marked
**N-Z Data Sets**.
- If the first letter of your LAST name is A through M, choose the data sets that are marked
- Download each of your four assigned data sets:
- If you are using the browser Chrome or the browser Firefox, click on the link for your data set, below. When prompted, save the file to your PSY-210_Summer2020_Unit09 folder.
- If you are using the browser Safari, right-click on the link for your data set, below, and select “Download Linked File.”
**A-M Data Sets**: Brewers’ 2019 Batting Averages, Women Actors’ Heights, NCAA 2020 Basketball Coaches’ Annual Compensation, Men Actors’ Age at First “Best Actor” Oscar Award**N-Z Data Sets**: Cubs’ 2019 Batting Averages, Men Actor’s Heights, NCAA 2019 Football Coaches’ Annual Compensation, Women Actors’ Age at First “Best Actor” Oscar Award- You’ll notice that each of these data sets is stored in a .csv file. As you remember, .csv stands for “comma-separated values,” and data stored in .csv can be imported into almost any data management platform.
- Therefore, third, import each of your four assigned data sets into a separate spreadsheet in your chosen data management platform following Andrews’ (2020) how-to article “
__How to Import Data from .csv Files__.”- After importing each of your four assigned data sets, save each of the four new spreadsheets naming each file,
**YourLastName**_PSY-210_Unit09_xxx, where xxx describes each of the four data set.
- After importing each of your four assigned data sets, save each of the four new spreadsheets naming each file,
- Fourth, calculate a percentile for each data value within each of the four data sets.
- Make sure you are calculating percentiles within each of the four data sets, not across the four data sets.
- After you have calculated a percentile for each data value within each of the four data sets, be sure to save each of your four spreadsheets.
- First, identity which four datasets you will be working with during this assignment by identifying the first letter of your LAST Name.
- Go to the
__Unit 9: Assignment #1 Discussion Board__and make a new post in which you do the following:- embed the screenshot of the percentiles you calculated during the tutorial (remember to embed and re-size according to the procedures of the Course How To).
- IF you were assigned the
**A-M Data****Sets**, write out and complete these sentences:- The Brewers baseball player with the highest 2019 batting average is ___ whose batting average was x.xxx, which is a percentile score of xx.xxx%.
- The shortest woman actor, among this sample, is ___ whose height is xx.xxx inches, which is a percentile score of xx.xxx%.
- The best paid NCAA basketball coach in 2020 is ___ whose annual compensation was $x,xxx,xxx, which is a percentile score of xx.xxx%.
- The oldest man actor to be awarded a “Best Actor” Oscar is ___ who received his Oscar at age xx.xxx years, which is a percentile score of xx.xxx%.
- IF you were assigned the
**N-Z Data****Sets**, write out and complete these sentences:- The Cubs baseball player with the highest 2019 batting average is ___ whose batting average was x.xxx, which is a percentile score of xx.xxx%.
- The shortest man actor, among this sample, is ___ whose height is xx.xxx inches, which is a percentile score of xx.xxx%.
- The second shortest man actor, among this sample, is ___ whose height is xx.xxx inches, which is a percentile score of xx.xxx%.
- The best paid NCAA football coach in 2020 is ___ whose annual compensation was $x,xxx,xxx, which is a percentile score of xx.xxx%.
- The oldest woman actor to be awarded a “Best Actor” Oscar is ___ who received her Oscar at age xx.xxx years, which is a percentile score of xx.xxx%.
- Disclaimers: We realize that some of the batting averages are artificially inflated because the players were rookies with late-season call-ups, and we apologize that the actor data accentuate a gender binary.
- In this assignment, you’re going to learn how to calculate
*z*-scores.- First, to learn what a
*z*-score is, read an excerpt from Investopedia’s (2020) article, “Z-Scores.” While reading this excerpt, make sure you understand the following:- A
*z-*score describes how far away a data value is from the mean, in terms of standard deviations. - A
*z*-score is 0.000 indicates that the data value is identical to the mean score. - A
*z*-score of 1.000 indicates that the data value is one standard deviation above the mean. - A
*z*-score of -1.000 indicates that the data value is one standard deviation below the mean
- A
- Second, read an excerpt from Wikipedia’s (2020) definition of “Z-Scores (Standard Scores).” While reading this excerpt, make sure you understand the following:
- A
*z-*score is how many standard deviations above or below the mean a data value lies in its distribution. - A
*z*-score is calculated by subtracting the mean and dividing the difference by the standard deviation. - Data values above the mean have positive
*z*-scores, while data values below the mean have negative*z*-scores.
- A
- Third, note that we usually italicize the
*z*in*z*-score. - Fourth, learn how to calculate
*z*-scores using your chosen data management system by completing Andrews’ (2020) tutorial “How To Calculate Z-Scores in Microsoft Excel, Google Sheets, or Apple Numbers.”- Again, to work with the “Sleep (in hours)” data, which you’ll need to work with for completing the tutorial, you should open a new spreadsheet and either
- type in the 18 values provided in Column A of Figure 1 of the tutorial
- OR
- copy those 18 values from YourLastName_PSY-210_Unit03_FrequencyDistribution_Andrews spreadsheet, which you created in Unit 3.
- Again, to work with the “Sleep (in hours)” data, which you’ll need to work with for completing the tutorial, you should open a new spreadsheet and either
- First, to learn what a
- After completing the tutorial to learn how to calculate
*z*-scores:- First, save the spreadsheet in which you calculated
*z*-scores during the tutorial with the filename YourLastName_PSY-210_Unit09_Sleep_Z-Scores - Second, take a screenshot of the z-scores you calculated during the tutorial.
- Name the screenshot YourLastName_PSY-210_Unit09_Z-Tutorial_Screenshot.xxx (where xxx is the filetype, for example, .jpg, .png, .jpeg and the like).
- Your screenshot should include only the part of your spreadsheet that contains data, NOT your entire screen.
- First, save the spreadsheet in which you calculated
- Now, let’s get some practice calculating
*z*-scores on other data sets.- First, identity which four datasets you will be working with during this assignment by identifying the first letter of your LAST Name.
- If the first letter of your LAST name is A through M, choose the data sets that are marked
**A-M Data Sets**. - If the first letter of your LAST name is N through Z, choose the data sets that are marked
**N-Z Data Sets**.
- If the first letter of your LAST name is A through M, choose the data sets that are marked
- Download each of your four assigned data sets:
- If you are using the browser Chrome or the browser Firefox, click on the link for your data set, below. When prompted, save the file to your PSY-210_Summer2020_Unit09 folder.
- If you are using the browser Safari, right-click on the link for your data set, below, and select “Download Linked File.”
**A-M Data Sets**: Cubs’ 2019 Batting Averages, Men Actor’s Heights, NCAA 2019 Football Coaches’ Annual Compensation, Women Actors’ Age at First “Best Actor” Oscar Award**N-Z Data Sets**: Brewers’ 2019 Batting Averages, Women Actors’ Heights, NCAA 2020 Basketball Coaches’ Annual Compensation, Men Actors’ Age at First “Best Actor” Oscar Award- You’ll notice that for this assignment, you have been assigned four different data sets than you were assigned in the previous assignment.
- Third, import each of your four assigned data sets into your chosen data management platform following Andrews’ (2020) how-to article “
__How to Import Data from .csv Files__.”- After importing each of your four assigned data sets, save each of the four new spreadsheets naming each file,
**YourLastName**_PSY-210_Unit09_xxx, where xxx describes each of the four data set.
- After importing each of your four assigned data sets, save each of the four new spreadsheets naming each file,
- Fourth, calculate a
*z*-score for each data value within each of the four data sets.- Make sure you are calculating
*z*-scores within each of the four data sets, not across the four data sets.
- Make sure you are calculating
- Fifth, after you’ve calculated a
*z*-score for each data value within each of the four data sets, be sure to save each of your four spreadsheets.
- First, identity which four datasets you will be working with during this assignment by identifying the first letter of your LAST Name.
- Go to the
__Unit 9: Assignment #2 Discussion Board__and make a new post in which you do the following:- First, embed the screenshot of the
*z*-scores you calculated during the tutorial (remember to embed and re-size according to the procedures of the Course How To). - IF, for this assignment, you were assigned the
**A-M Data****Sets**, write out and complete these sentences:- The Cubs baseball player with the highest 2019 batting average is ___ whose batting average was x.xxx, which is a z-score of x.xxx.
- The shortest man actor, among this sample, is ___ whose height is xx.xxx inches, which is a
*z*-score of x.xxx. - The second shortest man actor, among this sample, is ___ whose height is xx.xxx inches, which is a
*z*-score of x.xxx. - The best paid NCAA football coach in 2020 is ___ whose annual compensation was $x,xxx,xxx, which is a
*z*-score of x.xxx. - The oldest woman actor to be awarded a “Best Actor” Oscar is ___ who received her Oscar at age xx.xxx years, which is a
*z*-score of x.xxx.
- IF, for this assignment, you were assigned the
**N-Z Data****Sets**, write out and complete these sentences:- The Brewers baseball player with the highest 2019 batting average is ___ whose batting average was x.xxx, which is a
*z*-score of x.xxx. - The shortest woman actor, among this sample, is ___ whose height is xx.xxx inches, which is a
*z*-score of x.xxx. - The best paid NCAA basketball coach in 2020 is ___ whose annual compensation was $x,xxx,xxx, which is a
*z*-score of x.xxx. - The oldest man actor to be awarded a “Best Actor” Oscar is ___ who received his Oscar at age xx.xxx years, which is a
*z*-score of x.xxx.
- The Brewers baseball player with the highest 2019 batting average is ___ whose batting average was x.xxx, which is a
- First, embed the screenshot of the
- Now that you’ve had more experience with
*z*-scores, let’s explore the Empirical Rule.- First, to learn what the Empirical Rule is, read an excerpt from Wikipedia’s (2020) definition of the “Empirical Rule,” which is also known as the “68–95–99.7 Rule.” While reading this excerpt, make sure you understand the following:
- The Empirical Rule is also known as the 68–95–99.7 rule.
- 68–95–99.7 is a shorthand for remembering the percentage of values, in a normal distribution, that lie within certain bands around the mean.
- The bands refer to the prediction that
- plus or minus one standard deviation from the mean (which means plus or minus one
*z*-score) should contain about 68% of the distribution, - plus or minus two standard deviations from the mean (which means plus or minus two
*z*-scores) should contain about 95% of the data, and - plus or minus three standard deviations from the mean (which plus or minus three
*z*-scores) should contain about 99.7% of the data.
- plus or minus one standard deviation from the mean (which means plus or minus one
- Second, read an excerpt from Investopedia’s (2019) article, the “Empirical Rule.” While reading this excerpt, make sure you understand the following:
- about 68% of the data in a normal distribution will fall within ± 1 standard deviation from the mean, which means that about 68% of the data are likely to have a
*z*-score that is no greater than +1.000 and no less than -1.000; - about 95% of the data in a normal distribution will fall with ± 2 standard deviations from the mean, which means that about 95% of the data values are likely to have a
*z*-score that is no greater than +2.000 and no less than -2.000; and - about 99.7% of the data in a normal distribution will fall within ± 3 standard deviations from the mean, which means that about 99.7% of the data values are likely to have a
*z*-score no greater than +3.000 and no less than -3.000.
- about 68% of the data in a normal distribution will fall within ± 1 standard deviation from the mean, which means that about 68% of the data are likely to have a
- Third, to cement your learning about the Empirical Rule, look at Gernsbacher’s (2020) “Empirical Rule” handout. By this point, you should be familiar with the Empirical Rule, but this handout will help you visualize it.
- First, to learn what the Empirical Rule is, read an excerpt from Wikipedia’s (2020) definition of the “Empirical Rule,” which is also known as the “68–95–99.7 Rule.” While reading this excerpt, make sure you understand the following:
- Returning to the four data sets you worked with in Unit 9: Assignment #1, do the following with each data set.
- First, calculate a
*z*-score for each data value within each of the four data sets.- Make sure you’re calculating
*z*-scores within each of the four data sets, not across the four data sets. - Make sure you’re using the four data sets you worked with in Unit 9: Assignment #1, for which you previously calculated percentiles, but you did not previously calculate
*z*-scores on these four data sets. - After you’ve calculated a
*z*-score for each data value within each of the four data sets, be sure to save each of your four spreadsheets. - If you can’t remember how to calculate z-scores, review Unit 9: Assignment #2.
- Make sure you’re calculating
- Second, create a Cumulative Frequency Distribution Table for continuous data of the
*z*-scores in each of these four data sets. (You’ll use these Cumulative Frequency Distribution Tables to assess the Empirical Rule in each of the four data sets.)- If you need to brush up on creating Cumulative Frequency Distribution Tables for continuous data, look back at Unit 3: Assignment #4.
- In the Cumulative Frequency Distribution Table you create for each data set, use the following intervals (for each data set):
- -1.000 (Min) and -0.001 (Max)
- 0.000 (Min) and 1.000 (Max)
- -2.000 (Min) and -1.001 (Max)
- 1.001 (Min) and 2.000 (Max)
- -3.000 (Min) and -2.001 (Max)
- 2.001 (Min) and 3.000 (Max)
- -4.000 (Min) and -3.001 (Max)
- 3.001 (Min) and 4.000 (Max)
- -5.000 (Min) and -4.001 (Max)
- 4.001 (Min) and 5.000 (Max)
- Each of your four Cumulative Frequency Distribution Tables (one table for each data set) should look something like this.
- First, calculate a
- Next, assess the Empirical Rule in each of the four data sets by doing the following:
- First, because frequencies are always reported in proportions (e.g., 0.722) but for the Empirical Rule we want to report percentages (e.g., 72.222%):
- Use your chosen data management platform to convert each value in the Cumulative Relative Frequency column from a proportion to a percentage.
- If you can’t remember how to convert a proportion to a percentage, review pages 3 and 4 of Andrews’ (2020) tutorial “How To Calculate Percentiles in Microsoft Excel, Google Sheets, or Apple Numbers.”
- After converting each Cumulative Relative Frequency to a percentage, each of your Cumulative Relative Percentages should look something like Column L in this screenshot.
- Second, because assessing the Empirical Rule requires that we know what percentage of our data have a
*z*-score- between -1.000 and 1.000
- between -2.000 and 2.000
- between -3.000 and 3.000.
- However, because each of our intervals was only half that size, we need to identify the Combined Cumulative Relative Percentages that are illustrated in the right-hand margin of this screenshot.
- AND because the Empirical Rule only addresses the percentage of
*z*-scores that are between -3.000 and 3.000, when we assess the Empirical Rule, we need to ignore the percentage of*z*-scores greater than 3.000 or less than -3.000.
- First, because frequencies are always reported in proportions (e.g., 0.722) but for the Empirical Rule we want to report percentages (e.g., 72.222%):
- Go to the
__Unit 9: Assignment #3 Discussion Board__and make a new post in which you do the following:- IF, for this assignment (as well as for Unit 9: Assignment #1), you worked with the
**A-M Data****Sets**, write out and complete these sentences:- For the Brewers’ 2019 batting averages, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores. - For the women actors’ heights, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores. - For the NCAA 2020 basketball coaches’ annual compensation, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores. - For the men actors’ age at first Oscar award, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores.
- For the Brewers’ 2019 batting averages, xx.xxx% of the data set fell within ± 1.000
- IF, for this assignment (as well as for Unit 9: Assignment #1), you worked with the
**N-Z Data****Sets**, write out and complete these sentences:- For the Cubs’ 2019 batting averages, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores. - For the men actors’ heights, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores. - For the NCAA 2019 football coaches’ annual compensation, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores. - For the women actors’ age at first Oscar award, xx.xxx% of the data set fell within ± 1.000
*z*-score, xx.xxx% of the data set fell within ± 2.000*z*-scores, and xx.xxx% of the data set fell within ± 3.000*z*-scores.
- For the Cubs’ 2019 batting averages, xx.xxx% of the data set fell within ± 1.000
- IF, for this assignment (as well as for Unit 9: Assignment #1), you worked with the
- The last topic we’re going to cover in this Unit is the distinction between practical and statistical significance and how to quantify practical significance:
- First, to become acquainted with the distinction between statistical significance and practical significance, read an except from Frost’s (no date) blog post, “Practical vs. Statistical Significance.” While reading this excerpt, make sure you understand:
- Null Hypothesis Significance Testing (NHST) can tell us whether an effect is statistically significant, but our expertise and effect-size quantifications can tell us whether an effect is practically significant.
- Statistical significance refers to null hypothesis testing, but practical significance refers to the magnitude of the effect.
- Second, to become more familiar with practical significance, read Simply Psychology’s (2019) article, “What Does Effect Size Tell You?” While reading this excerpt, make sure you understand:
- “Effect size is a quantitative measure of the study’s effect. The larger the effect size, the more powerful the study.”
- Quantifying the size of an effect “promotes a more scientific approach” than null hypothesis significance testing because, “unlike significance tests, effect size is independent of sample size.”
- Cohen’s
*d*is a widely used measure of effect size. - A Cohen’s
*d*of 0.200 is considered a ‘small’ effect size, a Cohen’s*d*of 0.500 is considered a ‘medium’ effect size, a Cohen’s*d*of 0.800 is considered a ‘large’ effect size, and a Cohen’s d of 1.300 is considered a ‘very large’ effect size.
- First, to become acquainted with the distinction between statistical significance and practical significance, read an except from Frost’s (no date) blog post, “Practical vs. Statistical Significance.” While reading this excerpt, make sure you understand:
- Next, to learn how to calculate Cohen’s
*d*:- First, read an excerpt from Penn State’s (no date) lesson, “Practical Significance.” While reading this excerpt, make sure you understand:
- “Practical significance refers to the magnitude of the difference, which is known as the effect size. Results are practically significant when the difference is large enough to be meaningful in real life.”
- Cohen’s
*d*calculates the difference between two observed samples’ means and places that difference in standard deviation units by using a pooled standard deviation. - A Cohen’s
*d*of- 0.000 to 0.199 is considered little or no effect;
- 0.200 to 0.499 is considered a small effect size;
- 0.500 to 0.799 is considered a medium effect size;
- 0.800 to 1.399 is considered a large effect size; and
- 1.400 and above is considered a very large effect size.
- Second, read Poldrack’s (2020) Chapter 10, “Quantifying Effects.” While reading this excerpt, make sure you understand:
- “Cohen’s
*d*is used to quantify the difference between two means, in terms of their [pooled] standard deviation.” - Although the formula for subtracting one mean from another is simple, the formula for calculating the pooled standard deviation is more complex.
- Even for large differences (such as the difference between women’s and men’s heights), the two distributions (women’s heights and men’s heights) still overlap considerably.
- “Cohen’s
- First, read an excerpt from Penn State’s (no date) lesson, “Practical Significance.” While reading this excerpt, make sure you understand:
- Now, get some practice calculating Cohen’s
*d*:- Using your developing Google skills (remember that data scientists frequently use Google to learn how to do things), Google to find an online Cohen’s
*d*calculator.- You want to find a Cohen’s
*d*calculator for which you can input sample means and sample standard deviations (rather than input other measures). - Here are some example Cohen’s
*d*calculators (any one of which you can use — or you can find another!):- Social Scientists’ Effect Size Calculator
- Free Statistics’ Effect Size (Cohen’s
*d*) Calculator - University of Colorado’s Effect Size Calculator
- You want to find a Cohen’s
- Calculate Cohen’s
*d*for the following pairs of data sets:- Brewers’ 2019 Batting Averages versus Cubs’ 2019 Batting Averages
- Men Actors’ Heights versus Women Actors’ Heights
- NCAA 2019 Football Coaches’ Annual Compensation versus NCAA 2020 Basketball Coaches’ Annual Compensation
- Men Actors’ Age at First “Best Actor” Oscar Award versus Women Actors’ Age at First “Best Actor” Oscar Award
- You should already have the means and standard deviations for each of these eight data sets from the calculations you previously made in this Unit.
- When you are calculating each Cohen’s
*d,*be sure to maintain the order in which each member of a pair is listed above. - For example, when calculating the Cohen’s
*d*for Brewers’ versus Cubs’ batting averages, consider the Brewers’ batting averages your first sample (because they are listed first), and consider the Cubs’ batting averages as your second sample (because they are listed second). - Although Cohen’s
*d*values (like correlation coefficients and*z*-scores) can be negative as well as positive, for simplicity we’ve arranged each pair above in the order we have so that its Cohen’s*d*value will be positive. - Remember that good scientific practice requires that we always report three decimal places (if the Cohen’s
*d*calculator you use doesn’t provide three decimal places, you can add zeros for the missing decimal places).
- When you are calculating each Cohen’s
- Using your developing Google skills (remember that data scientists frequently use Google to learn how to do things), Google to find an online Cohen’s
- To follow Poldrack’s (2020) advice and investigate the amount of overlap between the two data distributions in the pairs of data sets listed above:
- First, go to Magnusson’s (no date) “Interpreting Cohen’s
*d*Effect Size: An Interactive Visualization,” which is an online visualizer. - Second, experiment with the visualizer by setting different Cohen’s
*d*values, which you can do either by sliding the dot on the left-right slider OR by clicking on the up-down vertical control as illustrated in this handout. - Third, notice that as the Cohen’s
*d*value gets larger, the two distributions get farther away; as the Cohen’s*d*value gets smaller, the two distributions get closer together, until they overlap completely. - Fourth, set the visualizer at a Cohen’s
*d*value that, prior to beginning this course, your intuitions would suggest would be a statistically significant effect.- Take a screenshot, and name the screenshot
**YourLastName**_PSY-210_Unit09_Intuition_StatisticalSignificant_Screenshot.xxx
- Take a screenshot, and name the screenshot
- Fifth, set the visualizer at a Cohen’s
*d*value that your current statistical thinking would suggest would be a*practically*significant effect.- Take a screenshot, and name the screenshot
**YourLastName**_PSY-210_Unit09_Cohensd_PracticalSignificant_Screenshot.xxx
- Take a screenshot, and name the screenshot
- Sixth, for each of the four Cohen’s
*d*s you calculated, set the visualizer at the Cohen’s*d*value you calculated.- Take a screenshot of each of the four visualizations, set at each of the four Cohen’s
*d*values you calculated, and name each screenshot**YourLastName**_PSY-210_Unit09_Cohensd_YYY_Screenshot.xxx, where YYY describes the data set pair you compared for that Cohen’s*d*.
- Take a screenshot of each of the four visualizations, set at each of the four Cohen’s
- First, go to Magnusson’s (no date) “Interpreting Cohen’s
- Go to the
__Unit 9: Assignment #4 Discussion Board__and make a new post in which you do the following:- First, report the Cohen’s
*d*you calculated for comparing Brewers’ 2019 Batting Averages versus Cubs’ 2019 Batting Averages**and**embed the screenshot you saved of that visualization. - Second, report the Cohen’s
*d*you calculated for comparing Men Actors’ Heights versus Women Actors’ Heights**and**embed the screenshot you saved of that visualization. - Third, report the Cohen’s
*d*you calculated for comparing NCAA 2019 Football Coaches’ Annual Compensation versus NCAA 2020 Basketball Coaches’ Annual Compensation**and**embed the screenshot you saved of that visualization. - Fourth, report the Cohen’s
*d*you calculated for comparing Men Actors’ Age at First “Best Actor” Oscar Award versus Women Actors’ Age at First “Best Actor” Oscar Award**and**embed the screenshot you saved of that visualization. - Fifth, embed the screenshot you saved of the visualization that, prior to beginning this course, your intuitions would suggest would be a statistically significant effect.
- Explain in one sentence or two why that visualization reflects your previous intuition.
- Sixth, embed the screenshot you saved of the visualization that your current statistical thinking would suggest would be a practically significant effect.
- Explain in one sentence or two why that visualization reflects your current statistical thinking.
- First, report the Cohen’s
- Meet online with your NEW Chat Group (which you formed during Unit 8) for a one-hour text-based Group Chat at a time/date that your Chat Group previously arranged.
**BEFORE MEETING WITH YOUR CHAT GROUP**do the following:- First, look at this sign that actually appeared at a skating rink.
- Try to think of any way that the second sentence makes sense. (Spoiler Alert: There isn’t any way that the second sentence makes sense.)
- Second, skim read Silver and McCann’s (2014) article, “How to Tell Someoneʼs Age When All You Know Is Their Name.”
- Notice the use of percentiles in each of the charts.
- Third, read a brief overview of the infamous Facebook-tinkered-with-our-emotions study, which was conducted in 2014.
- Look at the bar graphs of their data.
- Notice that the y-axes greatly exaggerate the differences (which, as you remember, is a no-no in graphing data).
- Therefore, it might not be surprising to learn that some of the Cohen’s
*d*s (effect sizes) in that study were as teeny tiny as 0.001 (although the results were statistically significant).
- First, look at this sign that actually appeared at a skating rink.
**BEGIN YOUR ONE-HOUR GROUP CHAT**by doing the following:- First, each Chat Group member should introduce themselves to the other members using the preferred first names that they provided back in Unit 1: Assignment #2.
- Second, to the degree that each Chat Group member feels comfortable, they should discuss with the other members their beginning and now mid-term level of statistics anxiety.
- Third, each Chat Group member needs to sum three two-digit numbers of their birthdate.
- For example, if you were born on March 18, 1999, your three two-digit-birthdate numbers are 03 (March), 18 (18th), 99 (1999), and the sum of your three two-digit birthdate numbers is 120 (03+18+99).
**DURING YOUR ONE-HOUR GROUP CHAT**do the following:- First, the Chat Group member with the
**highest**birthdate sum should lead a discussion on percentiles.- Do you find percentiles hard to understand? If so, why?
- Do you find percentiles easy to interpret? If so, why?
- Have you used percentiles before? If so, when?
- Were you able to interpret the percentiles in Silver and McCann’s (2014) article, “How to Tell Someoneʼs Age When All You Know Is Their Name.”
- How do you think you’ll use percentiles in the future?
- Second, the Chat Group member with the
**lowest**birthdate sum should lead a discussion on*z*-scores.- Do you find
*z*-scores hard to understand? If so, why? - Do you find
*z*-scores easy to interpret? If so, why? - Have you used
*z*-scores before? If so, when? - Were you able to think of any z-score that would make the second sentence in the skating rink sign make sense? (Spoiler Alert: Probably not.)
- How do you think you’ll use
*z*-scores in the future?
- Do you find
- Third, the Chat Group member with the
**neither the highest nor the lowest**birthdate sum should lead a discussion on the Empirical Rule. For Chat Groups with only two members chatting, they should share leading this discussion.- Do you find the Empirical Rule hard to understand? If so, why?
- Do you find the Empirical Rule easy to interpret? If so, why?
- How do you think you’ll use the Empirical Rule in the future?
- Fourth, the Chat Group member with the
**highest**birthdate sum should lead a discussion on effect size (and Cohen’s*d*).- Do you find effect size and calculating Cohen’s
*d*hard to understand? If so, why? - Do you find effect size and calculating Cohen’s
*d*easy to interpret? If so, why? - Were you surprised to learn that the Cohen’s d (effect size) in the infamous Facebook-tinkered-with-our-emotions study, was so teeny tiny?
- What factors could have led to such a teeny tiny Cohen’s
*d*from data that were nonetheless statistically significant? (Hint: Sample size) - How do you think you’ll use effect size and calculating Cohen’s
*d*in the future?
- Do you find effect size and calculating Cohen’s
- Fifth, the Chat Group member with the
**lowest**birthdate sum should lead a discussion on the distinction between practical versus statistical significance.- Do you find distinguishing between practical versus statistical significance hard to understand? If so, why?
- Do you find distinguishing between practical versus statistical significance easy to interpret? If so, why?
- How can we apply the distinction between practical versus statistical significance to understanding the infamous Facebook-tinkered-with-our-emotions study?
- How do you think you’ll distinguish between practical versus statistical significance in the future?
- First, the Chat Group member with the
**AT THE END OF YOUR ONE-HOUR GROUP CHAT**do the following:
- Nominate one member of your Chat Group (who participated in the Chat) to make a post on the Unit 9: Assignment #5 Discussion Board that summarizes your Group Chat in at least 200 words.
- Nominate a second member of your Chat Group (who participated in the Group Chat using the browser Chrome on their laptop, rather than on their mobile device) to save the Chat transcript, as described in the Course How To (under the topic, “How To Save and Attach a Chat Transcript”).
- This member of the Chat Group needs to make a post on the Unit 9: Assignment #5 Discussion Board and attach the Chat transcript, saved as a PDF, to that Discussion Board post.
- Remember to attach the Chat transcript by clicking on the word “Attach.” (Do not click on the sidebar menu “Files.”)
- This member of the Chat Group needs to make a post on the Unit 9: Assignment #5 Discussion Board and attach the Chat transcript, saved as a PDF, to that Discussion Board post.
- Nominate a third member of your Chat Group (who also participated in the Chat) to make another post on the Unit 9: Assignment #5 Discussion Board that states the name of your Chat Group, the names of the Chat Group members who participated the Chat, the date of your Chat, and the start and stop time of your Group Chat.
- If only two students participated in the Group Chat, then one of those two students needs to do two of the above three tasks.
- Before ending the Group Chat, arrange the date and time for the Group Chat you will need to hold during the next Unit (Unit 10: Assignment #5).
- Record a typical Unit entry in your own Course Journal for the current Unit, Unit 9.
Congratulations, you have finished Unit 9! Onward to |