Who, What, When, Where, and How?
Who, What, When, Where, and How?
When designing a study, it’s important to understand what or who you’re studying, when you’re measuring it (and how often), and where you’ll be collecting your data. These decisions will determine other aspects of your study design, such as your sample and method of data collection.
Unit of Analysis
The unit of analysis is a key part of any research study—it tells us what or who we’re studying. Are we looking at individuals, like one person’s experiences or behaviors? Or are we focused on groups, like families, schools, communities, or entire countries? Getting clear about the unit of analysis helps make sure your research question, data, and analysis all line up.
Sometimes the unit of analysis is obvious. For example, if you’re studying how students cope with stress, you’re probably focusing on individual students. But other times it’s more complicated. For example, if you’re comparing how different schools support student mental health, then your unit of analysis isn’t the student—it’s the school.
Why does this matter? Because the unit of analysis affects how you collect and interpret your data. If your unit is the individual but you accidentally collect data at the group level—or vice versa—you might draw the wrong conclusions.
Here’s an example from family science: let’s say you want to study how conflict affects marital satisfaction. If your unit of analysis is the individual, you might look at each partner’s perspective separately. But if your unit is the couple, you’d need to combine or compare both partners’ responses to make conclusions about the relationship as a whole. Your findings and interpretation could be very different depending on that choice.
Time Scale
Time is an important aspect of research. When are we measuring what we’re measuring, how often will we measure it, and are measuring something happening right (prospective) or something that happened in the past (retrospective)? Two common types of research are cross-sectional and longitudinal. Cross-sectional studies are studies that take place at one point in time only. Longitudinal studies are those that take place at multiple time points. This might be by getting data from the same respondents at multiple times (before and after an event, for example, or once a month for a year), which is called a panel study, or it could be by getting data from different respondents but at multiple times (repeating a survey every year with different samples each time, for example), which is called a repeated cross-section.
The choice of time matters greatly in research. To understand why, let’s first look at an example that is familiar to many family scientists: marital satisfaction. This has obviously been an important topic to family scientists for a long time. Early studies of marital satisfaction showed a pattern – satisfaction was high in early years, declined over time until about a decade in, stayed low for a few years, then started to increase again. This curvilinear relationship came to be known as the U-shaped curve and was an accepted fact by many marriage researchers (CITE Booth). However, not all family scientists
Prospective vs. Retrospective Research
For Rachel A.’s memory: https://research.ebsco.com/linkprocessor/plink?id=fc34e209-bea8-3d2a-8a56-f9c4f77009e2
https://research.ebsco.com/linkprocessor/plink?id=792f1873-8bb5-345e-b07d-7eee3b14bd53
Another aspect of time that is important to consider is whether you’ll be collecting prospective or retrospective data. That is, are you collecting data as it is happening (prospective) or after it’s happened (retrospective)? A lot of studies are, by necessity, retrospective. Retrospective research is a research method that examines past events to gain a deeper understanding of current outcomes. It utilizes existing data, records, or participants’ memories to explore the relationships between factors and outcomes that have already occurred. If you’re being asked to report on your childhood experiences as a young adult, you’re giving a retrospective account of the experiences. Sometimes, the experience may not be particularly long ago but can still be retrospective – asking a couple to recount how they met, even if that was last year, is still retrospective. You can probably already guess what the risks are with retrospective data: memory! Human memories are inherently fallible. If you’re asking someone to report on something in specific detail (like how many hours they spent vacuuming or doing other household chores last week, as an example), you may or may not get the story of exactly how things went.
* This image was created using napkin.ai; however, the concept, design direction, and creative vision were conceived by Dr. Knight
(If OTL could figure it out, I would love for these figures to sit right next to each other)
In contrast, measuring things as they happen is prospective. Sometimes, this actually looks like data being collected during an experience (like in an experience-sampling study, where someone is asked to enter data onto a device – like an app on their phone – at planned or random times throughout the day), or it can also mean collecting data throughout a longer period that then, when everything is said and done, happens to contain the experience of interest (like a study that starts on children in first grade and collects measures of their family life and mental health throughout their school years – at the end of high school, the researchers could look back and see which children experienced parental divorce and when, then examine how their mental health changed at those times). If we take our example above of housework, asking someone to record what they’re doing for each 15-minute period of a day for two days (this is a version of a time diary) will give very different data than the retrospective survey above. In fact, this exact comparison, of survey to time diary, produced some very important and impactful findings regarding household labor distribution across the transition to parenthood in a 2015 study called “The Production of Inequality: The Gender Division of Labor Across the Transition to Parenthood.”
The choice of prospective vs. retrospective data is sometimes made for a researcher – if an event has already taken place, they can’t go back in time to set up the study beforehand! However, if there is an option to use or collect data as something is happening, a researcher should balance what may be gained vs. what resources would be required. Sometimes retrospective data is actually completely sufficient for a study’s purposes!
Environment
Once you’ve decided who and what you’re analyzing and when, you must consider where your study will take place, also known as the research environment. “Environment” could mean the physical space in which the research is done (do you meet your interview participants at their home? A coffee shop? Your on-campus lab?) or even an online space, if you’re collecting data via an online survey or remote interview. Online studies are only growing, of course, but special considerations must be made for online research, such as the inability to explain the informed consent to participants – it must be clear enough for them to be able to read and understand it themselves. If your research will take place in a physical setting, being thoughtful about access (this relates to the ethical principle of justice), comfort, power, and convenience will be important. Consider if a setting is realistic, but also how controllable it is. If you need a lot of control for a study (you need each person’s experience to be exactly the same, like in an experiment), then having the study take place in a lab might be important. If you don’t care as much about control, or have a way to control the setting elsewhere, having a study team actually go to a participants’ home may be the most accessible method. Safety must also be considered though – there’s a lot that goes into research decisions!
Research Methods
Now that you know who and what you’re studying and when and where your study will take place, it’s time to determine how your study will take place, including the specific techniques and procedures you will use to carry out your study. These include decisions about sampling, variables and measurement, data collection, and data analysis. Research methods should align with your research question, objectives, and methodology. Some researchers consider things like the unit of analysis, time scale, and methodology part of their research design. However, in this book we will use the term research methods to refer to the following:
Sampling
A sampling strategy is how researchers recruit participants for their study. Ideally, the sample should represent the bigger group they’re interested in—like choosing college students from different majors if studying student stress. Sometimes researchers use random sampling, while other times they use whoever is available (called convenience sampling). The method they choose depends on their goals and resources. What matters most is being clear about who was included and why, so others can understand how well the findings might apply to the larger population.
Variables and Measurement
Variables are the things researchers are trying to study—like stress levels, test scores, or physical activity. Measurement is how they figure out the amount or presence of these variables. A good measurement tool gives results that are both accurate (measuring what it’s supposed to) and reliable (producing similar results each time). Sometimes, measurements are objective (like heart rate), and sometimes they’re based on opinions or feelings (like survey answers). It’s important that researchers pick tools that match their research goals and make sure they use them correctly, so the data they collect is solid and trustworthy.
Data Collection
Data collection is how researchers gather the information they need. Different methods are used depending on the kind of question being asked. For example, interviews are great for deep, personal insights; surveys work well for reaching lots of people quickly; observations are useful for watching behaviors in real settings; and biological measures (like blood pressure or cortisol levels) are used when researchers need physical data. Each method has its pros and cons, and choosing the right one helps researchers get the kind of data that best fits their study.
Data Analysis
Data analysis is how researchers make sense of the information they’ve collected. In studies with numbers (quantitative), they might use statistics to look for patterns, compare groups, or test whether something caused a change. In studies with words and experiences (qualitative), they might read through interview transcripts and look for common themes or ideas. Sometimes, researchers use a mix of both. The goal is to find meaning in the data and answer the research question in a clear and honest way.
