Data Services Case Studies
==========================

The Data Services team has assisted students and researchers on a wide
variety of topics. Below, you can seea sampling of projects and software 
we have helped with.

Text Analysis using R
---------------------

|text analysis 1|

.. collapse:: More info

    .. container::

        This student wanted to study how cultural and national 
        narratives influence individual and collective consciousness. 
        She wanted to explore how Chinese citizens built their new 
        identities and discuss their past after Mao. Text analysis of 
        tens of thousands of film reviews on Douban, a popular Chinese 
        online database. 
        
        We showed her how she could use R to analyze those reviews and 
        visualize the results. She found that there are four camps of 
        mentalities or ways of perceiving the national past, 
        (progressivism, fatalism, optimism, and anomie) that are very 
        distant from each other. 

        |text analysis 2|

        Zhang, J. (2024). *How We Reconcile with the Past: National 
        Narrative and Individual Consciousness* [Brandeis University]. 
        https://doi.org/10.48617/etd.1131

        https://jiayizhangg.shinyapps.io/farewell/#topic=0&lambda=1&term=

Data Analysis with Stata
------------------------

|stata 1|

.. collapse:: More info

    .. container::

        This student wanted to study the the impact of higher education on 
        the political participation and engagement of female students of 
        color. She took data from the Political Engagement Project (PEP), 
        2003-2005 [United States] (ICPSR 36977) and analyzed it with Stata. 
        We helped her with data cleaning, data analysis, and data 
        presentation. She found that education serves as a direct influence 
        for building skills, knowledge, engagement, and momentum to 
        participate and engage in politics. 

        |stata 2|

        Shealy, J M. (2023). *The Influence of Higher Education on the 
        Political Engagement and Participation on Women Students of 
        Color* [Brandeis University]. 

Designing Experiments in PsychoPy
---------------------------------

|psychopy 1|

.. collapse:: More info

    .. container:: 

        |psychopy 1|

        In Psychology, designing experiments that participants can use 
        on any device is a powerful way to make life easier for your 
        participants. A common tool for this is PsychoPy, an experiment 
        builder that lets you use a graphical interface to define the 
        broad strokes of the experiment and Python code for specifics. 
        Experiments can be hosted for free on pavlovia.org. We can help 
        students put together the experiment they want to in any part of 
        it. 

        |psychopy 2|

Image Processing in Python 
--------------------------

|image recognition 1| |image recognition 2|

.. collapse:: More info

    .. container::

        This student wanted to take images they had captured of moths 
        over the course of a summer and extract moths of a certain size 
        from the image to determine the percent of the board covered 
        with moths and to attempt to group them together by shape size 
        and color. We helped to automate this process in Python for 
        hundreds of images. 
        
        This was a rare instance where Data Services generated the 
        majority of the code for this project, as the student had no 
        coding experience, and we had the bandwidth to help make a pilot 
        for the project.  The moth images were fed into a clustering 
        model, and this model was able to differentiate taxonomic 
        groupings based on their size and coloration. The student also 
        found that they could use the percent of the board covered by 
        moths as proxy for moth biomass. 

        Shimazu, F. Y. (2023). *Reducing the Participation Barrier to 
        Insect Monitoring Through the Image Processing Program: 
        Measuring Insect Board Coverage Using Adaptive Thresholding 
        (MIBCAT)* [Brandeis University]. 
        https://doi.org/10.48617/etd.1074


.. |text analysis 1| image:: /_static/images/case-studies/text_analysis1.png

.. |text analysis 2| image:: /_static/images/case-studies/text_analysis2.png

.. |stata 1| image:: /_static/images/case-studies/stata1.png

.. |stata 2| image:: /_static/images/case-studies/stata2.png

.. |psychopy 1| image:: /_static/images/case-studies/psychopy1.png

.. |psychopy 2| image:: /_static/images/case-studies/psychopy2.png

.. |image recognition 1| image:: /_static/images/case-studies/image_recognition1.png
    :width: 49.5%

.. |image recognition 2| image:: /_static/images/case-studies/image_recognition2.png
    :width: 49.5%