3. The Role of Theory in Research
3.2. Levels of Analysis
Victor Tan Chen; Gabriela León-Pérez; Julie Honnold; and Volkan Aytar
Learning Objective
Define levels of analysis and how they influence our approach toward research.
In the hard sciences, researchers can study phenomena at very different levels—from stars and planets, to land masses and biomes, to microscopic cells and microbes, to the fundamental particles that constitute our material reality. In a similar way, sociologists study the social world using different levels of analysis (Figure 3.1). The three broad levels of analysis are micro, meso, and macro. We can think of this as a continuum from the lowest levels of analysis to the highest—from interactions between individuals, to comparisons of entire societies (Video 3.1).
At the micro level, sociologists examine the smallest levels of interaction—in some cases, just “the self” alone. Micro-level analyses might include one-on-one interactions between couples or friends. Or, perhaps a sociologist is interested in how a person’s perception of self is influenced by their social context. In each of these cases, the level of inquiry is micro.
In addition to deciding on a level of analysis, sociologists also need to choose the units of analysis they wish to study (we’ll discuss units of analysis in depth in Chapter 6: Sampling). A unit of analysis is what we are really studying—the class of phenomenon we ultimately wish to comprehend. Typical units of analysis include individuals, groups, organizations, and countries. The unit of analysis at the micro level is often the individual, but that’s not always the case. For example, when studying social interactions, we might focus not on individuals per se, but on dyads, pairs of people who are connected through that interaction (see the sidebar Sociology across Boundaries to learn about research on interracial couples). Or, we might focus on the interactions themselves—that is, making every conversation into a unit of analysis. This brings up another feature of units of analysis: they don’t have to be people. We can analyze texts like books and magazines, visual content like paintings and TV shows, or events like protests and strikes (we’ll talk more about some of these possibilities in Chapter 15: Materials-Based Methods). Our choice of a unit of analysis will depend on what we truly want to understand—individuals, or conversations, or books, or whatever.
When sociologists investigate groups, their inquiry is at the meso level. Sociologists who conduct meso-level research might study how norms of workplace behavior vary across professions (with specific occupations, like doctors or firefighters, being the unit of analysis), or how children’s sporting clubs are organized (with a club being the unit of analysis).
At the macro level, sociologists examine social structures and institutions. Research at the macro level examines large-scale patterns. For instance, sociologists have long studied the process and impacts of globalization, the growing interconnectedness of the economies and cultures of the world. A research project that examined how nations are connected economically or culturally would be an example of a macro-level study. In this case, the unit of analysis would be a country, though sociologists can also design macro-level studies where the unit of analysis is larger—say, a region of the world—or smaller—say, cities or states in a particular country.
Video 3.1. Macro and Micro Perspectives. This video illustrates the differences between micro- and macro-level approaches.
Note that the boundaries between micro, meso, and macro levels of analysis are relative and somewhat arbitrary. For example, some international nongovernmental organizations, like the International Monetary Fund or United Nations, represent many countries: when we study them, are we operating at a meso or macro level? Likewise, a quantitative analysis of poverty using state-level data might be considered a “macro” approach relative to ethnographic observations in poor neighborhoods. Using these labels helps us situate our research alongside other work, but we don’t want to get hung up over the precise categories.
Let’s take a closer look at some specific examples of sociological research to better understand each of the three levels of inquiry described previously. Some topics are best suited to be examined at one level, while other topics can be studied at each of the three different levels. The particular level of inquiry might shape a sociologist’s questions about the topic. Sociologists might also view the topic from different angles depending on the level of analysis being employed.
Work by Stephen Marks offers an excellent example of research at the micro level. In one study, Marks and Shelley MacDermid (1996) draw from prior micro-level theories to empirically study how people balance their various life roles—for instance, their identities and activities as parents, partners, and workers. In this study, the researchers found that people who experience balance across their multiple roles report lower levels of depression and higher levels of self-esteem and well-being than their less “balanced” counterparts. In another study, Marks and his colleagues (2001) examined the conditions under which spouses feel the most balance across their roles. For women, having more paid work hours and more couple time were crucial. For men, having leisure time with their families was important, but their “role balance” decreased as their work hours increased.
At the meso level, sociologists tend to study the experiences of groups or the interactions between groups. For example, Jason Spicer and Christa Lee-Chuvala (2021) studied a global network of banks dedicated to upholding social and environmental values. Tracking how membership in the network changed over time, they found that credit unions and other banks owned by their customers rather than investors were more likely to remain within the network and stay committed to their values. In a much different study of group-level data, Michael Messner (2009) conducted research on children’s sports leagues. Messner studied interactions among parent volunteers, among youth participants, and between league organizers and parents. He found that the adults who run such leagues structure the children’s gameplay in ways that reinforce gender boundaries and hierarchies.
Sociologists who conduct macro-level research study interactions at the broadest level, such as comparisons across cities or nations. For example, David Frank, Bayliss Camp, and Steven Boutcher (2010) examined changes in laws regulating sex across a number of countries between 1945 and 2005. They found that laws regulating rape, adultery, sodomy, and child sexual abuse shifted in focus from protecting families to protecting individuals. In another macro-level study, Leah Ruppanner (2010) studied how national levels of gender equality in 25 different countries affected how couples divvy up housework. Ruppanner found that as women’s representation in a country’s parliament increased, so, too, did how much time men in that country participated in housework. (Note that we don’t always know what is driving the changes and associations we observe across these sorts of studies—in this case, how exactly greater women’s representation in a country’s government is linked to greater male participation in housework, or whether the first really causes the second. We’ll have more to say about these issues later in this chapter.)
Let’s consider one important sociological topic—poverty—and then consider how researchers study it from each of the levels of inquiry just described. At the micro level, they might study what everyday life is like for people who are homeless by observing them where they panhandle and sleep. One example of a micro-level study of extreme poverty that uses this approach is Gwendolyn Dordick’s book Something Left to Lose (1997), based on the author’s intimate observations of the lives of homeless people in New York City. At the meso level, sociologists often study poverty by looking at issues that arise within poor neighborhoods—as Robert Sampson has done intensively in the city of Chicago, as described in his book The Great American City (2013). And at the macro level, they might compare poverty across communities within a nation—or even compare rates of poverty by country. David Brady and Amie Bostic (2015) used survey data from 37 countries to test earlier theories about the mix of social policies that were most effective in reducing poverty rates across those countries. They found that policies providing generous government assistance that was more universal—that is, not targeted just at the poor—were associated with the sharpest reductions in poverty.
While it is true that some topics lend themselves to a particular level of inquiry, many social phenomena can be studied from any of the three levels. The choice depends on the specific interests of the researcher, the approach they would like to take, and the sorts of questions they want to be able to answer about the topic.
Sociology across Boundaries: A Q&A with Chinyere Osuji
Chinyere Osuji is a sociologist who studies how Africans and members of the African diaspora understand racial and ethnic “others.” She examines the meanings they give to ethnic and racial differences as well as how they use other social categories to reinforce, blur, or erase those boundaries. Her first book, Boundaries of Love (2019), compared how Black-White couples in Brazil and the United States make sense of race. Currently, she is working on two book projects: one about African immigrants confronting race in the U.S. nursing profession, and another on anti-Blackness and White supremacy in South Korean media.
What drew you to study Brazil and to do your fieldwork in that country?
In graduate school, I found out that Brazil had the largest number of people of African descent in the world outside of Nigeria. I’d already had an interest in race in Latin America, and given the millions of people who identify as negro, Brazil became an obvious choice when trying to understand people of the Afro-Atlantic. Since I already spoke Spanish, I figured Portuguese would be easy to pick up.
In Boundaries of Love, you discuss the fluid and complex ways that people think about their racial and ethnic identities—not just contrasting the U.S. and Brazilian understandings of racial identity, but also talking about how multiple aspects of a person’s identity and social context shape their personal experience of race. Can you talk about how your own understanding of race and ethnicity has evolved over the years, and how you came to the nuanced theoretical perspective you put forward in your book?
I grew up in Chicago in the 1980s. My parents didn’t understand U.S. race relations. They’d come here in the late ’70s from Nigeria with the idea that they were going to get a college education and then return home to the strong economy and newly minted democracy they’d left behind. But while they were living abroad, Western-backed coups and economic policies destabilized Nigeria and many other countries in the global south. Staying in the United States seemed like the best option.
Today, bilingualism is highly valued. However, when my parents arrived, the United States was informally an English-only society. My teachers encouraged my parents not to speak to me or my siblings in their native tongue, Igbo. They thought I would have trouble with English, even though my parents speak both languages fluently.
Our family lived on the north side of the city, where many recent immigrants from Africa tend to live. However, because we lived in public housing, we were surrounded by all sorts of low-income families, including Black Americans, who were the largest group in our building. There were also other families of African immigrants, Puerto Ricans, Mexicans, and even Eastern European refugees living there. Since Africans were their racial status quo, my parents thought everyone with light skin—whether Puerto Rican, Middle Eastern, or Chinese—were oyibo: White people. Some Black American neighbors complained about our “stinky” African food. Meanwhile, African immigrants tsk-tsked the single-headed households of our Black and Puerto Rican low-income neighbors. One Afro-Cuban neighbor looked like us, but since she spoke only Spanish, l found her confusing—all the other foreign-born Blacks I had encountered spoke English.
Middle-class Black families lived in distant neighborhoods like Hyde Park or the south suburbs, so my understanding of Blackness from kindergarten to eighth grade was limited. It was only when I attended Whitney Young High School—also Michelle Obama’s alma mater—that I saw Black Americans across class backgrounds. I learned that many behaviors that Americans—both Black and non-Black—had insisted to me were “Black” were really traits of the urban poor.
In college, I studied abroad in the southern Spanish city of Granada, where the old Moorish palace the Alhambra still stands. In my classes, I learned that Africans ruled over much of the Mediterranean long before the concept of race existed—sometimes in cooperation with Jewish and Christian brethren, sometimes not. Several people I met in Granada looked like what we might call “mixed” or “biracial” in the United States. There, I learned to speak Castilian Spanish, lisp and all. When native Spaniards tried to speak ill of me—thinking I was an African street vendor who could not understand them and deserved poor treatment—I replied in their own tongue. In Granada, I also witnessed discrimination against Spain’s Roma population. Despite the fact that gitanos were seen as thieves and ne’er-do-wells, gitano culture was nationalized to the point that flamenco was seen as Spanish, not as a cultural product of their disenfranchised Roma people. Living in Spain—the birthplace of modern racism 550 years ago—made me interested in how people around the world understand ethnic and racial boundaries.
In your book, you push back against notions in popular culture that Brazil is a racially harmonious utopia. How did your work lead you to this view?
I came across several articles by Edward Telles, a future advisor and mentor of mine at UCLA. He had written a lot about racial equality in Brazil. According to his research, Afro-Brazilians earned less than their White peers even when they worked in the same occupations. They had lower levels of education even when they had parents of the same socioeconomic background. They also lived in neighborhoods that were racially segregated—even if less so than hyper-segregated U.S. cities. These racial inequities were not fully explained by class in the ways that many Brazilians had assumed.
When I first went to Brazil for my research, I lived in a wealthy White neighborhood in São Paulo. I soon realized that everywhere around me were White people who looked Italian or Portuguese. The only Black or brown-skinned people there worked as domestics, doormen, or day laborers. My landlady was strange, saying things questioning my intelligence or belittling me as a human. I found it odd how her olive-skinned daughter, who was close in age to me, avoided me. The only saving grace was my White roommate, Celda, who was really kind. She had grown up in one of Portugal’s colonies in Africa, but African independence had forced her Portuguese colonizer parents to leave when she was young. Over drinks one night, Celda explained that our landlady had told her secretly that she did not like Black people and had advised her daughter not to spend time with me outside of the home. That was my cue to leave. At the end of the month, Celda helped me pack up all of my belongings. In the morning, I told our landlady that I did not like how she had been treating me and that I had decided to move out. She was completely gobsmacked as I left the keys on the table. This housing situation was when I first confronted the fraud of Brazilian racial harmony.
Your interviews with interracial couples about their past and present “romantic careers” delve into intimate territory: sexual preferences, racial and gender stereotypes, fraught family dynamics. How did you approach your interviews on these topics, and how did you guard against the possibility of alienating or offending your interviewees?
There were a lot of uncomfortable moments as I was doing those interviews. I reminded myself that I was just listening to their stories. No matter how uncomfortable or unrelatable they could be, they were still their stories and deserved all the respect and care I could muster. Any analysis I did would not occur in the moment. I would take notes on my emotions on the ride back to my apartment in Rio de Janeiro or Los Angeles. If necessary, I would discuss interesting encounters with my roommate or my friends.
Through my interviews with couples, I learned that men often expect women to do the emotional labor of recounting details of their relationship. It didn’t matter whether they were Black or White, or living in Los Angeles or Rio de Janeiro: wives crafted the narrative of their relationships, and husbands rubber-stamped them. Men’s voices are important, too, and I wanted to give them space to share their own experiences of their interracial relationships outside of their wives’ interpretations. For this reason, I conducted the rest of my interviews one-on-one—making sure that the spouse was not within hearing distance. This was more of a struggle in lower-income neighborhoods, where there was not a lot of space for interviewees to be isolated from other family members. I will never forget going to a shantytown to interview a man about his previous relationships. He quickly glanced from side to side and spoke in a hushed tone, looking very, very uncomfortable. The walls in his home were thin, and since we could hear his wife moving about in the next room, she could probably hear us as well.
What advice would you give an aspiring sociologist if they wanted to pursue the sort of cross-national interview-based research you do?
Do not do it until after you have finished your dissertation. My original goal was to study only Brazil, but I had a fellowship that required that I stay in Los Angeles for one more year. I thought, “Why not make this a comparative project?” I have no regrets, but I wouldn’t recommend it to a graduate student. I’d recommend it to somebody who’s already done work in one society or can collaborate with other researchers.
Key Takeaways
- Sociological research can occur at micro, meso, or macro levels of analysis, ranging from the study of individuals and individual-level interactions, to the exploration of groups and organizations, to the analysis of entire countries and societies.
- Some topics lend themselves to one particular analytical level, while others could be studied from any, or all, of the three levels of analysis.
Exercise
Look over the examples of research studies discussed in this section and see if you can figure out what the unit of analysis was. (You can go off the description in the next, or download the actual article if you have library access, as described in Chapter 5: Research Design.) For some of these studies, the unit of analysis will be obvious, but for others, you might make a case for one or the other—or even multiple units of analysis.
The different levels of aggregation at which social scientists can study phenomena, which can range from the macro (societies) to the micro (individuals).
The lowest level of analysis, which in the social sciences typically involves the study of individuals.
The class of phenomena (e.g., individuals, groups, objects, societies) that researchers want to learn about through their research.
Pairs of people who are connected through social interaction.
The middle level of analysis, which in the social sciences typically involves the study of organizations or other kinds of groups.
The highest level of analysis, which in the social sciences often involves the study of communities, societies, or countries.