Big Data In Human Services Discussion

Big Data In Human Services Discussion
Introduction
Welcome back to your virtual internship at the Riverbend Community Action Center! So far, you have been introduced to your overarching project of using data analytics to help RCAC evaluate the effectiveness of their Ruby Lake Teen Homelessness Task Force and figured out how to store the data using the appropriate data models. Now it’s time to learn how to actually analyze the data that’s been stored.
It’s time for another meeting with your mentor, Brenda.
Mentor Talk
Riverbend City Community Action Center: Mentor’s Office
Check in with with your mentor, Brenda.
All right, welcome back! I hope you’re finding our talks useful. Today’s session definitely should be… we’re going to be talking nuts and bolts here.
At this point, we’ve got the data stored in an organized way so that we can find what we need. Now it’s time to do some actual data analysis using statistics. The results of the statistical analysis are what we need to understand what’s happening, based on the data, and make some recommendations.
Today we’re going to walk through how to do one statistical analysis, an Independent Samples T-Test, using MicroSoft Excel. You can do these types of analyses with programs like SPSS and SAS, but since we already invested in buying the MS Office suite, and Excel can run these calculations too, we just use Excel. Using Excel takes some practice, but keep at it! It’s not that complicated once you get used to the program.
You know one of the main reasons teens can end up homeless is because they’re having a lot of problems with their parents. So when the teens first come to us for services, one of the measures we give them in their intake packet is a Parental Rejection Scale. The final score on the scale is a number that can range from 0 (no sense of rejection from parents) to 100 (total rejection by parents). This gives us a sense of how much of an issue this may be for the teenager.
I’m sure you also know that there’s research out there that says LGBT teens are more likely to run away from home—or get kicked out of their homes—because they feel rejected by their parents. We want to see if that’s the case here at the RCAC or not. So we randomly selected the Parental Rejection Scale scores of 30 LGBT teens and 30 heterosexual teens who’ve come for services over the past year.
Remember, an Independent Samples T-Test compares the means of two different groups. Here, we have two independent groups: one group of LGBT teens and one group of heterosexual teens. We have the score for each group. Once we tell Excel to run the t-test, it will do all the rest for us.
You can see that there are two columns here: one labeled LGBT and one labeled Heterosexual. In each column you can see there are lists of numbers. Those numbers are each teen’s score on the Parental Rejection Scale. So the first LGBT teen provided a score of 79; 79 out of 100 suggests this teen felt pretty rejected by their parent. For the first Heterosexual teen, that person reported a score of 76; 76 out of 100 suggests that teen also felt pretty rejected by their parent.
Looking down the columns of scores, they don’t look all that different, do they? Well, this is why we do statistical analyses: we don’t know if there’s a statistically significant difference until we do the analysis! We can’t just trust our first impressions.
Now, we’re going to start telling Excel to run an Independent Samples T-Test. Remember that you don’t have to calculate the means of the group yourself: Excel will do all of that. All you have to give Excel is the raw data (in this case, the scores on the test) in the correct columns.
Here’s something to remember: if you tried opening up your Excel and don’t have the Data Analysis options available, then you need to install the Analysis ToolPak. This is available as a free add-in in Excel: I’ll give you some instructions later on how to do that.
Now, let’s run this t-test!
Go up to the top menu and click on “Data.”
Look to the left under the Data menu and click on Data Analysis.
You’ll see that we have two options for an Independent Samples T-Test:
· t-test assuming equal variances, and
· t-test assuming unequal variances.
We have to know up front if the variances in our two groups (LGBT teens and heterosexual teens) are the same or different. In other words, is the range in scores (in other words, the variance) for the LGBT teens similar to the range in scores of the heterosexual teens or not?
Luckily, Excel will calculate that for us, too!
If you scroll up through the list of data analyses, there’s an option for “F-test Two Sample for Variances.” Select that option and click OK.
A new box comes up, asking you for input. Variable 1 range is going to be the scores for the LGBT teens. You don’t have to input each and every value for all thirty teens! Instead, just highlight the first value in the LGBT column (row