CS + X

The focus this week was helping the students create a presentation for the celebratory banquet at the end of their program, which was a great opportunity for the students to reflect on the work that they have done. The presentation is meant to be a comprehensive explanation of the topics we covered in class, all the personnel involved in this program and classes, and their thoughts and reflections on their experience. I designed the outline for the students, so they were able to spend time writing and reflecting. In addition, they also spent a considerable amount of time on the design and background images on their slides, which made me smile because I enjoyed watching them explore the themes and create their own to best represent themselves.

On Tuesday, we watched a video on the history of linguistics in order for the students to have a more firm undertanding of the field. We found a great video through TedTalk that detailed the history of linguistics and addressed the historical prejudice and discrimination that played a role in spoken language. The video used examples from all over the world about government leaders celebrating one dialect, or variation, of a language and shaming other varities in the hopes of uniting the country with a shared language. The impacts of those decisions are still felt today. While the students had not necessarily heard about the historical decisions and impacts on language, they have learned about those figures and they have experienced hardships in their own life, so it was a poignant moment.

Unfortunately, there was inclement weather on Thursday, so we were not able to see the students for their last session with us! We were not able to finish the presentation nor give them a proper send-off, so there are plans in the works in order to see them once again.

For the group of local high school students, we met with them to check-in to see their progress on their group project. They were able to add a lot more determiners, nouns, adjectives, and intransitive verbs to their sentence programs. They were curious to address a problem about the program emitting more than one adjective in the noun phrase, and even though in English we can have a lot of adjectives in a row, it is not very common to have 10 in a row like this program was doing! We discussed in the code that the program recursively calls itself if the random number generator generates a 1, if it is a 2, the program ends and compiles a noun phrase. There is a possibility that the program will keep receiving 1’s as input and generate some unwiedly long sentences. So the students brainstormed that they can change the program to include a for loop instead of recursion to have more control.

Another problem was that the program would select contradictory adjectives like “The big little camel eats”, so we also brainstormed on ways to help the program understand that those words are contradictory. The students thought of separating lists into “good”, “bad”, and “neutral” adjectives or attribute based lists like “small” or “large”, and they settled on first trying to limit the amount of adjectives picked, and then deciding on this area later.

Study Skills and Belonging

I spent more time redesigning the data parser that I made in the last week. With help from friends, we found a new way of looking at the data with the literal_eval() function that was able to recognize that the data was a dictionary inside a string, and that saved so much time and lines of code. The simplified results of this code is beautiful, and it looks like JSON, which is a lot easier to work with. I am excited to be able to perform the sentiment analysis this week because the data is so much easier to work with.

This week we met to discuss ways of cross-sectioning the data in our analytical and descriptive analyses. To refer back to our main goal, we are hoping to identify trends of successful study behavior among students with different learning styles in order to give meaningful advice to incoming students in this discrete math and computer science course. We used a latent class analysis to identify four classes of students with similar self-reported learning styles and ideas about growth mindset and learning, and we found some weak correlations between how they spent their time. To give an example, class 1 which earned the highest grades spent an hour more a week with practice examlets than class 2, but it was not statistically significant.

This was disheartening because we were not finding strong trends in behavior that could help us, so we thought about sectioning the data on new indicators to see if we could find more significant results. We discussed using hometown zipcodes to see if there were trends there so we split the data into zipcodes for domestic students and continents for international students. We found that international students spent a significantly longer time studying than domestic students, but we found it difficult to draw meaningful insights from this. Were international students spending more time studying because of the unfamiliarity of the content? Or because English isn’t their first language? It is not appropriate to make these large assumptions when we do not have the data to back it up.

So in the later half of the week we pivoted back to our original latent class identified groups and we began to compare them to each other in more of a categorical sense:

• Class 1 - Growth Mindset and Value (Good Performers)
• Class 4 - Growth Mindset and lesser value
• Class 2 - Better grades than class 3 even though no growth mindset
• Class 3 - Less value less growth mindset less belonging and lower grades

I am curious what other factors are driving student grades, I think there is a social element and resource that we will glean from the interview data that can better explain why student with no or less growth mindset earn on average higher grades than a group with higher growth mindset.

General Reflections

I am sad to see this group of Upward Bound students “graduate” because it felt like we did not have enough time together! We were forming great relationships and engaging with the material, and now they are off to learn more about other subject areas. I wish them the best, and hopefully we will see them briefly in the future to finish up their presentation.

For the next group, I plan on scaling back the Introduction to Python Google Colab worksheet because I would like to readjust the scope of the course. The course is not an introduction to python programming, the course is about engaging with computer science and linguistics in a relevant way. The students will need to know what variables and functions are, but we are able to learn about those in a condensed way as to share information, but not overwhelm.

This week I will code more interviews and perform the sentiment analysis on the qualitative interview data. I am hoping to find trends on what sentiment and polarity numbers occur with which tagged keywords as well as trying to quantify something so subjective like emotions.