This post explains how the research behind South Asian American Experiences in Schools was designed, including the methodological decisions that shaped the findings discussed elsewhere on this site.
When I wrote South Asian American Experiences in Schools: Brown Voices from the Classroom, I was trying to position it both as an academic work that would have value for education practitioners but also as something non-education people might find useful. But the book, at its core, was rooted in academic work. I wish the research behind it had been straightforward: a question, a study, results, conclusions. But in reality it was actually a winding and stuttering path that led to an awareness of a gap that made me go “wait, why isn’t anyone studying this more directly?”
When I was beginning this research (especially around 2013-2014), there was already extensive research on Asian American students. But studies specifically on South Asian Americans were very limited.
And since “The Asian American Experience” isn’t one big monolith, this can be an issue. We (meaning different Asian subgroups, communities, castes, cultures, religions, and all the weird intersections of all these things) have distinct histories, relationships to immigration, values and beliefs, and even face different stereotypes and challenges. So when data isn’t disaggregated to look at distinct subgroups, even one as large as “South Asian Americans,” on their own, it becomes harder to talk precisely about what’s happening in schools and what those students experience.
So I wanted to look into it. (And I did so so knowing even that looking at “South Asian American experiences” was still creating the same problem, but just reducing its scale. But you have to start somewhere).
At the same time, there’s a world of literature on multicultural education. So I did what a lot of researchers do at the beginning: I read EVERYTHING I could.
I read everything I could find on Asian American education research AND everything I could find on multicultural education research. And I began to see certain patterns: some frameworks, some debates, some recurring themes. And yet few of the studies really looked at how South Asian Americans were experiencing schools. What their experiences were like with their teachers, whether they felt supported in schools, what school was like for them.
And then I started moving toward building a way to study this.
Before I surveyed students, I ran a small scale needs assessment study with teachers. I wasn’t looking to do a definitive study, but to figure out what beliefs and assumptions might be floating around in the first place, especially related to stereotypes like the model minority myth.
Even with a small sample, the sample of teachers was effective in guiding what I should examine on the student side, and what kinds of teacher beliefs and assumptions were likely shaping student experiences.
In addition to the initial needs assessment survey, I also conducted interviews with four teachers to dig deeper into what the needs assessment was hinting at.
I structured that interview using Stanford University’s design school’s Design Thinking model. This wasn’t a design project, but I liked Design Thinking here because it’s built for complex, human-centered problems; we needed to be able to talk about culture, bias, and student experience, so I wanted a human experience centered method that could help surface the things people might not otherwise articulate.
Together, the needs assessment and teacher interviews informed the development of the research instrument that would eventually provide the data at the heart of this book.
The full survey was designed to explore South Asian Americans’ experiences of being students—especially in the everyday context of school: teacher relationships, classroom expectations, stereotypes, assumptions, and support (or lack of support).
I wanted to look at a few intertwined layers:
Structurally, the survey included 24 quantitative (Likert-style) questions and three qualitative open-ended narrative questions.
The quantitative items allowed for measurement of how strongly participants agreed or disagreed with statements about their experiences in areas like support, teacher assumptions, classroom dynamics, and perceived understanding.
The three open-ended, constructed-response items generated qualitative responses, which I later coded into categories for additional analysis. These were meant to capture participants’ language, tone, and meaning-making (things that don’t fit neatly in multiple-choice questions).
I didn’t want the qualitative part to just become a list of unrelated anecdotes; I wanted to be able to identify if/when patterns emerged in the responses. So I translated narrative patterns into simple coded categories to see whether the qualitative responses lined up with quantitative patterns.
Here’s how the coding worked:
Constructed response #1 (could teachers have supported you more?)
Constructed response #2 (did teachers understand you and your background?)
Constructed response #3 (overall experience in context of teacher relationships)
This gave me a way to look for relationships between the scaled survey items and the themes in participants’ narratives.
I was going to be writing about racism, cultural proficiency, and people’s lived experiences in schools, which are loaded subjects.
I wanted to be sure that the instrument was clear, accurate, and valid; this is why the process of developing and testing the instrument involved multiple stages, including pre-testing items, refining wording, and running cognitive interviews. I’m making note of some of the process here in this post for anyone curious about the rigor in the survey instrument design.
Because the survey asked participants to reflect on potentially sensitive experiences involving race, culture, and teacher relationships, I was careful in designing the response scales themselves. Each response option included its own descriptor rather than relying only on end-point anchors, a choice intended to reduce ambiguity for respondents. The primary survey used a four-point agreement scale:
I did NOT include a neutral midpoint. One consideration in scale design is that neutral options can allow respondents to avoid making a directional choice. For this instrument, I used a four-point scale that required participants to indicate whether they agreed or disagreed (while still allowing for strength of agreement).
Although some literature recommends a greater number of response options, I determined that a four-point scale best fit the structure and purpose of this instrument.
For selected items, the order in which response options were presented was reversed as a strategy intended to reduce patterned responding.
These decisions were not arbitrary; they were grounded in existing survey design research and were made to ensure the instrument was as clear, valid, and reliable as possible.
I conducted a structured pre-testing process to evaluate whether the items were clearly measuring distinct constructs. I used a modified version of a two-stage sorting procedure based on the work of Moore and Benbasat (1991). (The procedure is described in depth in an Appendix of my book.)
After the two-stage sorting process, I also ran an additional round of pretesting using cognitive interviews with participants who fit the profile of the intended study population. (This procedure is also described in depth in an Appendix of my book.)
The survey instrument was designed for participants who self-identified as South Asian backgrounds. Eligibility was confirmed at the start of the survey; respondents who did not identify accordingly were screened out from results. Participants were asked open-ended demographic questions about their cultural background, allowing them to describe ethnicity, nationality, race, religion, or other relevant identifiers in their own words. This approach was intentional, as fixed categories can obscure nuance within South Asian communities.
Additional demographic questions captured participants’ highest level of education and the type of K through 12 school they attended (public, private, charter, etc), in order to contextualize responses across different educational settings. These questions were designed to provide meaningful context while preserving participants’ ability to self-describe rather than be confined to predetermined identity labels.
The survey results are the main subject of my book. And they aren’t entirely pleasant. But I do want to share that the book itself isn’t meant to demonize or demoralize teachers (a group I was a part of before, during, and after this research, until 2020), although the findings might make us feel that way.
The point of the work was to look honestly at what South Asian American students experienced in schools, and at how teacher beliefs, assumptions, and cultural preparedness shaped those experiences.
The process of designing the study itself was a critical part of making sure the resulting findings would be a valuable contribution to the literature. And I think they are.
