2. Data Analysis Milestone Four – In Milestone Three, you described the steps to perform the statistical test. In Milestone Four, you will perform the statistical test using StatCrunch to answer your research question IHP 525 Milestone Four Guidelines and Rubric. You must follow these steps to successfully complete the assignment: IHP 525 Milestone Four Steps.pdf
Here are some FAQs for your reference:
1. How to run the t-test for independent samples in StatCrunch, considering the categorical variable? (to determine differences between the groups)
In StatCrunch, you will click on Stat, T Stats, Two Samples, With Data. You have to "build" the where statements using the categorical variable to identify each group. For example, the where statements to determine age differences in the use of EMR to send records to a third party (RecordsOnline_Send3rdParty) among individuals with depression (MedConditions_Depression) are:
- Sample 1 "RecordsOnline_Send3rdParty" = 1 and MedConditions_Depression = 1
- Sample 2 "RecordsOnline_Send3rdParty" = 2 and MedConditions_Depression = 1
Remember that your interpretation of the p-value and conclusion in terms of the research question is what matters! ;)
IMPORTANT: Three variables describe the use of EMRs. You must perform the t-test for independent samples for each EMR variable.
2. When is it OK to remove outliers?
Outliers should not be removed just because they are extreme values. We need to consider whether those values are telling us something else that can explain our results. To determine outliers in our dataset, we create a range using Q1 and Q3. Any numbers outside the range are considered outliers: [Q1 - 1.5(IQR), Q3 + 1.5(IQR)]
To determine anomalies (more extreme outliers): [Q1 - 3(IQR), Q3 + 1.5(IQR)]
However, these ranges simply determine outliers, which you already know from the box plots!
To determine whether to remove an outlier from our analysis, we need to "learn" about the specific data point: Was it an error in data entry? Was this individual really a part of the population of interest? Or is it a person who was very sick and had other comorbidities or complications?
Unfortunately, our data set is very limited. My suggestion is to include the outliers in the statistical test and explain the presence of outliers as a limitation and the need for additional variables to better explain the association between the two variables of interest.
3. What are parametric vs nonparametric statistical tests?
Parametric statistical tests are used when the data are normally distributed in which the mean and standard deviation are appropriate summary statistics; whereas nonparametric statistical tests are used when the data are skewed in which we use the median and IQR to summarize the data. StatCrunch has some nonparametric statistical tests under Stat (Nonparametrics). You can find a little more information about each nonparametric test in the Article Review Sample.
4. Final Project Templates: Article Review and Data Analysis
- Do not refer to the articles by the title, for example, The article "Title of the article" by "author's last name" .... As you can imagine this is an example of wordiness. You should simply use an in-text citation, for example, Poggi (2024) ...
- Please remember that there are two different final projects in this course: (a) article review and (b) data analysis. The article reviewis about the two articles you selected from the list of two different options, and you have developed through the article review assignments. The data analysis has been developed through milestone assignments using the data set provided to you.
- For your convenience, I have color-coded each section using the milestone and article review assignment prompts. Hope this is helpful when combining the assignments into your final submission2. Data Analysis Milestone Four – In Milestone Three, you described the steps to perform the statistical test. In Milestone Four, you will perform the statistical test using StatCrunch to answer your research question IHP 525 Milestone Four Guidelines and Rubric. You must follow these steps to successfully complete the assignment: IHP 525 Milestone Four Steps.pdf
Here are some FAQs for your reference:
1. How to run the t-test for independent samples in StatCrunch, considering the categorical variable? (to determine differences between the groups)In StatCrunch, you will click on Stat, T Stats, Two Samples, With Data. You have to "build" the where statements using the categorical variable to identify each group. For example, the where statements to determine age differences in the use of EMR to send records to a third party (RecordsOnline_Send3rdParty) among individuals with depression (MedConditions_Depression) are:- Sample 1 "RecordsOnline_Send3rdParty" = 1 and MedConditions_Depression = 1
- Sample 2 "RecordsOnline_Send3rdParty" = 2 and MedConditions_Depression = 1
[Q1 - 1.5(IQR), Q3 + 1.5(IQR)]To determine anomalies (more extreme outliers):
[Q1 - 3(IQR), Q3 + 1.5(IQR)]However, these ranges simply determine outliers, which you already know from the box plots!To determine whether to remove an outlier from our analysis, we need to "learn" about the specific data point: Was it an error in data entry? Was this individual really a part of the population of interest? Or is it a person who was very sick and had other comorbidities or complications?Unfortunately, our data set is very limited. My suggestion is to include the outliers in the statistical test and explain the presence of outliers as a limitation and the need for additional variables to better explain the association between the two variables of interest.3. What are parametric vs nonparametric statistical tests?Parametric statistical tests are used when the data are normally distributed in which the mean and standard deviation are appropriate summary statistics; whereas nonparametric statistical tests are used when the data are skewed in which we use the median and IQR to summarize the data. StatCrunch has some nonparametric statistical tests under Stat (Nonparametrics). You can find a little more information about each nonparametric test in the Article Review Sample. 4. Final Project Templates: Article Review and Data Analysis- Do not refer to the articles by the title, for example, The article "Title of the article" by "author's last name" .... As you can imagine this is an example of wordiness. You should simply use an in-text citation, for example, Poggi (2024) ...
- Please remember that there are two different final projects in this course: (a) article review and (b) data analysis. The article reviewis about the two articles you selected from the list of two different options, and you have developed through the article review assignments. The data analysis has been developed through milestone assignments using the data set provided to you.
- For your convenience, I have color-coded each section using the milestone and article review assignment prompts. Hope this is helpful when combining the assignments into your final submission.