Hello, I'm Karen Monsen, professor at the University of Minnesota School of Nursing. This is the second of five data applications modules for our social determinants of health, data to action specialization. In this module, we will explain and conduct t-tests and use box plots to visually display our findings. These learning activities provide context and information for our hands-on analysis. In this course, we focus on comparing two groups, males and females, using inferential statistics to inform our understanding of social determinants of health for these two groups. In each data to action module, we'll consider ways to ensure we are ethical analysts and allies to the most vulnerable. The scientific community has discovered many nuances regarding the spectrum of genders. It is generally accepted that a person's internal sense of gender identity may or may not be the same as the sex assigned at birth. In many ways, structural discrimination by race and ethnicity is similar to that by gender. We are committed to centering our data analysis around equity in gender representation. Recognition of the diversity of gender has led to re-examination of how the concepts of sex, gender identity, and sexual orientation are measured. Better measurement will improve the ability to identify sexual and gender minority populations and understand the challenges they face, such as disparate and inequitable treatment, including harassment, discrimination, and violence, which in turn affects health and social outcomes. Though knowledge of these disparities has increased significantly over the past decade, glaring gaps remain often driven by the lack of reliable data. This issue is revealed in our data sources. We have male-female data and we're conducting comparisons using t-test for two groups. We present this as an opportunity to discuss different approaches to gender measurement and analysis by group. Consider the need for anonymity and data and also the problem of invisibility in data. These are complex problems. The solutions will differ depending on many factors. Remember our moral and ethical mandate to protect the most vulnerable groups, even as we discover critical information to improve population health. Prioritizing those who are marginalized as we strive to increase health equity and improve health for all. In this module, we'll be conducting a t-test. A t-test is a statistic that checks if two means or averages are reliably different from each other. As you watch this video, consider, we will be analyzing differences in a continuous variable for two groups in a data sample. In each of our two textbooks, we point you to optional readings pertaining to t-test analysis. Likewise, for each book, we point you to optional readings pertaining to box plots. Box plots are great because they make it easy to see distributions of a sample. Box plots show the median of the sample, its upper and lower 25th quartiles, and its minimum and maximum values. Outliers are also shown. This makes it easy to see differences side by side. As you will see in our analysis, we'll use R to create these plots for us and also to conduct our t-tests analysis. Just a quick reminder here that to perform a t-test, you need to have two groups and the continuous variable of interest. If the group means differ, the t-test tells you how significant the differences between the group means are. That is, how likely these differences in means could have happened by chance. Also, when we look at box plots, we're not looking at the means, we're looking at medians. This is a little detail that we should know and appreciate how it might make the pictures slightly different from the t-test findings. All this talk about significance reminds me to provide a resource about p values. One way to talk about statistical significance is to look at the probability that our results are within what would be expected in a normal distribution. Shorthand for this is called a p-value and we typically use 0.05 as our cutoff for significance. For more information, check out this video. Tying all of this into our course material. Recall the notion of collective impact that is critical to the success of a data to action initiative. Our collective impact example for this course is early wins in early childhood. As you read, think about the data sources that the pilot project used and how the team used data to tell the story as an inspiration and especially to decision-makers to create policy change. Recall also that we have positive system archetypes to guide us in our data to action efforts. For each collective impact example, we draw on a positive archetype that underlies that example successes. In this case, we see that early wins in early childhood project success leverage the collective agreement archetype. Recall that a number of these positive system archetypes can be used to overcome our classic mental models that create system failures and perpetuate negative health and social outcomes. Review Collective Agreement versus Tragedy of the Commons. As you review the positive archetypes, do you agree that collective agreement positive archetype was instrumental in the success of early wins in early childhood? Which other positive archetypes may have contributed to the success? With all of these exciting new ideas and building on our previous data experience, you're ready to move on to Course 2, Module 2, Part 2, t-tests analysis and visualization. In this case, we're going to be able to answer the question, are differences by gender likely due to chance?