Massive Dataset: Center for Disease Control and Prevention's National Health Interview Survey data from 2016 and 2014.

How did I Manipulate it? Python coding language.

What was the goal?

To highlight different parts of the data based on sexual orientation criteria. There is a section in the NHIS of self identified sexual orientation in order to narrow the vast amount of data included in the survey to just stuff related to this topic and specific questions related to these people i also included in my new file is sex, and racial identity.

I created a new database of just the sexual identity, sexual identity, and racial identity. From here , using the code found through our Github page (https://github.com/dlachenm/LGBTQ-class-project.git ), one can reproduce the code to add more columns to create information with that data. I wanted to create a baseline of information for people more experienced than me, or someone that has access to more accurate data to compare it to.

Conclusions:

When looking at the data and asking myself what could come of this project I realized that I truly did not feel comfortable drawing conclusions from it. This is government information related to personal emotions and identity. The American public has no reason to be honest in this space. The government could use this information in a plethora of different ways, and it serves as no distinctly positive thing for the average american to fill out honestly. Those in fear of repression or persecution may not answer honestly to these questions and there is no way of assessing the validity of the information.

Data biases:

In addition to this, data bias is real. The way that questions are asked has a very large impact on the answer one receives. If asked in a different frame, or with long answer instead of multiple choice, this survey could have gotten very different results. There are biases implicit in everything and that is not restricted to simply government data.

Where to go from here:

To compare this interview survey with a community interview survey given by community members to community members would be of great interest to me. To test my theory on the comfort level of individuals talking to the government and talking to their peers. One could also take the data for what it is and compare the upcoming survey data with that data developed from the 2017 survey and create a change over time.