Behavioral Science Tools

Introduction

Every profession has its tools. Carpenters swing hammers, farmers drive tractors, surgeons wield scalpels, and brewers tend fermenters. What tools do behavioral scientists use?

In this chapter, we will learn about the tools that make modern behavioral research possible. With these tools, researchers today collect, analyze, and interpret more data about the human experience than at any other point in history. And, as with Chapter 1, you are not going to just read about these tools—you are going to use them.

In Module 2.1, we will introduce you to online research by having you participate in several online studies. Participating in these studies will allow you to see how behavioral research operates and what it looks like from the participant's perspective. Afterward, we will explore the tools that make research possible (Figure 2.1). We will learn how researchers develop ideas and review the existing literature in Module 2.2. We will learn about designing studies and gathering data in Module 2.3, and we discuss how researchers analyze data in Module 2.4. Finally, the chapter ends by describing tools researchers use to communicate their results and share their data. By walking through each step of the research process, we will see how the results of one project often serve as the starting point for another, continuing the research cycle.

Throughout the chapter, we will use an example project to make each step concrete. The project examines personality traits that are associated with navigating life alone versus in a committed relationship. The project builds upon the personality research we learned about in Chapter 1 and provides the opportunity to see how researchers might investigate such a question. So, when you are ready, let's dive into the tools of behavioral research!

Chapter Outline

Introduction to Online Research

Behavioral research is about curiosity. Once researchers find a topic they are interested in, they observe people's behavior, form hypotheses, test their ideas in studies, and use data to draw conclusions. This process occurs in dozens of disciplines and across all kinds of organizations. It also has big consequences. Research helps doctors understand how to treat mental illness; it allows businesses to market products; it helps lawmakers understand crime; and it enables scholars to explain everything from decision-making to discrimination. In other words, curiosity can have a big impact.

Today, most behavioral research doesn't happen in labs or classrooms—it happens online. Researchers around the world rely on specialized tools to find participants, conduct studies, and collect data. While we will learn to use these tools throughout this book, for now, we will introduce online research by focusing on the role of research participants.

The Role of Participants in Science

Behind every discovery in behavioral science there are people who have chosen to provide data by participating in a research study. Historically, these people have been hard to find.

Before the internet, researchers often found participants from local universities or community advertisements. These methods were slow, and they limited who could participate to people physically near the researchers. Then, the advent of online tools transformed how behavioral scientists find participants. Today, digital platforms connect researchers with participants from all around the world. These websites allow researchers to post studies targeted at specific people. Participants browse projects by topic, time commitment, and compensation, completing the ones that look appealing.

Online research has become so common that more than 80% of published studies in social psychology have at least some participants from online sources (Zhou & Fischbach, 2016). When recruiting participants online, researchers can quickly reach people from different geographic locations, age groups, educational backgrounds, and life experiences. They can also get these people to do things that would be too burdensome offline, such as filling out a survey multiple days in a row.

Although online tools make research easier, it's important to remember that participants are more than data points. They are people who offer their time and attention so researchers can test ideas and build knowledge. Without participants there is no study, no analysis, and no scientific discovery.

In a few minutes, you will see how behavioral research operates online by participating in a few studies. You will create a participant account on a commonly used platform operated by CloudResearch, called Connect. Then, you will participate in at least three studies. These studies are posted by researchers from all around the world, spanning all the disciplines and organizations introduced in Chapter 1. After you complete a few studies, we will step into the shoes of a researcher and learn about the tools that made your experiences possible.

📊

Research Activity 2.1: Participating in Online Studies

You might be thinking: What will I learn from participating in studies? The answer is several things, including some you can't learn any other way.

First, you will learn about the incredible variety of questions behavioral scientists investigate. While you read about this diversity in Chapter 1, participating in studies will allow you to see it for yourself. In one study, you might help researchers understand how people make financial decisions, how people form first impressions, or how they solve complex problems. In another, you might participate in a psychology experiment, take a survey about your political views, or play an economic game. You might even be asked to serve as a juror in a mock trial. Each study will give you a different look at behavioral research, and unlike the demonstrations in Chapter 1, the data from these studies will help researchers better understand behavior.

Second, you will experience the tools and methods researchers use to collect data online. You might be asked to complete studies with interactive tasks, survey questions, or games and simulations. Many of these tasks will be presented on survey platforms that you will learn to use later in the book.

Finally, and perhaps most importantly, serving as a participant will make you a better researcher. You will learn what it feels like when instructions are clear (or confusing), when a study is engaging (or boring), and when the compensation seems fair (or not). These insights into what it's like to be a participant will be invaluable when we start designing studies and recruiting participants later in the book. In fact, one of the first steps many professional researchers take when getting started with online research is to participate in studies so they can understand the issues mentioned above.

Plus, did we mention you'll get paid? Most projects on Connect pay between $10 and $15 per hour, although the typical project only lasts about 10 minutes. Even so, when was the last time you had a class assignment that allowed you to earn money while learning?

Creating a Participant Account

To help set up your participant account, we have created a video: https://bit.ly/Ch2_Pacct. You can watch the video on YouTube or follow the instructions below. Signing up takes a few minutes and requires completing an onboarding process, but once it is done you will be ready to participate in studies.

To create a Connect account, navigate to the CloudResearch website: https://bit.ly/3RoVBnZ. Enter your name and college or university email address (e.g., a .edu address within the U.S.). Select the "Research in the Cloud" book for how you heard about the site. Then select "Create Account."

You will be asked to verify your email address. Click on a link sent to your email account to do this. After verifying, you can log back into the site.

When you log in, you will see a screen like Figure 2.2. This screen welcomes you and starts the onboarding process. Onboarding takes a few minutes, but it is an essential part of maintaining high data quality on Connect, which we will learn about in Chapter 10.

After onboarding, you will be granted access to the site. Before taking studies, you will be asked to provide some basic demographic information and configure a payout method (Figure 2.3). While most participants connect to a PayPal account, you can choose Amazon gift cards or a transfer directly to your bank account.

Participating in Projects

The participant dashboard looks like Figure 2.4. From the dashboard you can see available projects.

The projects on Connect are designed by researchers from all over the world. Many of these researchers work at colleges and universities, but others work at multinational corporations, government agencies, non-profits, or private businesses. Some are even student researchers just like you will be soon!

Using the menu on the left, you can review your project history, cash out your earnings, communicate with researchers, or complete your user profile in the "About You" section.

When you see an interesting project, select "View." This will take you to a page that summarizes what you will be asked to do and presents any special instructions for the project (Figure 2.5). Special instructions are where researchers communicate technical requirements (e.g., downloading software), warn about sensitive content, or inform participants that the project involves follow-up surveys.

To accept a project, click "Start." You will be presented with a link to the survey. Clicking "Go to Study" will take you out of Connect and to the study created by the researcher. Usually, this study is in Qualtrics, Survey Monkey, Engage, or a similar survey platform. Once you complete the project, you will be automatically directed back to Connect or given a completion code to copy and paste into the submission box on Connect.

And that's all there is to it! You are ready to participate in behavioral research. As a reminder, you should complete at least three projects. In these projects you can find everything from consumer surveys and experimental studies to mock jury trials, interviews with artificial intelligence, and fun personality tests. After you have taken a few studies, come back here to learn about the research process and how these studies got on Connect in the first place.

From Participant to Researcher

Now that you have taken part in research as a participant, it's time to switch perspectives. Let's step into the shoes of a behavioral researcher.

Behind each study you completed was a researcher, or team of researchers, who asked the question, designed the study, chose the tools, collected the data, and figured out what the results meant. As you transition from a participant to a researcher you should shift from thinking, "What is this study about?" to asking "How can I create this study myself?" Being a researcher means thinking critically about how to measure behavior, how to find good information, and how to tell a story with your results.

To get started, create a Connect researcher account. This will allow you to create studies and collect your own data down the road.

To create an account, go to: https://account.cloudresearch.com. Make sure you choose "Connect for Researchers" (Figure 2.6) and use the same email and password you used as a participant. Once you have logged in, you are officially a researcher on Connect. With this account, you will be able to create and launch studies like those you just participated in.

The researcher side of Connect looks like Figure 2.7. From this screen, you can use the icons on the left to communicate with participants, create studies, and manage participants.

If you click the "Create Project" button, you will enter the study set-up window. From there, you can prepare a project by describing what participants need to do, pasting a link to your survey, determining who is eligible based on hundreds of demographic characteristics, and making decisions about how many people you want to recruit. Once you launch the project, it is available to participants.

Throughout the rest of this book, we will show you how to set up online projects, recruit participants, and gather data that answers questions you are curious about. But first, we need to walk through the research process. This is where research gets exciting. So, let's begin.

The Research Cycle

As you participated in studies, perhaps you wondered: Who is conducting this research? How did these projects come to be? What happened before this study appeared online, and what will the researcher do with the data afterward?

These questions are about the research process, which begins long before participants see a study and continues after they have completed it. Let's outline this process before exploring the tools that make it possible.

Step 1: Generating ideas and reviewing the literature. Research begins with curiosity and a research question. But before designing a study, researchers need to know what others have already discovered about the topic. Specialized search tools like Google Scholar make it possible to quickly search the published literature in a scientific area. Reading this literature helps researchers develop questions and connect their ideas to existing theory.

Step 2: Designing studies and collecting data. Next, researchers must choose how to measure the things they are interested in. Will they use surveys? Cognitive tasks? Physiological measurements? Will the study take place online or in person? For many research questions, online platforms for designing research studies, like Qualtrics, Survey Monkey, and Engage provide a powerful and user-friendly way to design professional surveys.

Step 3: Finding participants. Once the study is designed, researchers must find the participants. Online platforms for finding participants like Connect make it possible to find diverse participants in hours rather than weeks or months.

Step 4: Analyzing data and visualizing the results. After collecting data, researchers use statistical tools to analyze patterns in the data and create visualizations that help communicate the findings. These tools transform raw numbers into insights and conclusions.

Step 5: Sharing findings. Finally, researchers share what they have learned. Once the findings of a study are disseminated to others through publications and presentations, they become part of the scientific record. As other scientists review the work and form new questions, the research cycle begins again.

In the rest of this chapter, we focus on the tools researchers use to develop ideas, design studies, analyze data, and share their results. As we just learned, the process begins with a literature review.

Where do research ideas come from? Curiosity is the simplest answer. Whenever you observe people around you, notice trends in society, or learn about behavior from books, movies, and media, you might think: why do people do that? Anytime you have this experience, you have an idea for research.

Let's imagine you are interested in the Big Five personality traits described in Chapter 1. Let's also imagine you notice surveys and other data showing an increasing number of people are moving through life without getting married. You decide to explore whether people who are single have different personality characteristics than those who are married, and whether one group of people is happier than the other.

Before you read further, look at the Big Five traits in Chapter 1 (see Box 1.1). Create some hypotheses about what the personality differences might be between people who stay single versus those who get married. Write these down in your portfolio.

The Value of a Literature Review

When researchers are curious about a question, one of the first things they do is explore what is already known about the topic.

Has anyone investigated this question? If so, what did they find? Perhaps previous research has only looked at the question among college students but not older adults (or the reverse). Understanding the research that has already been done allows scientists to design studies that build upon existing knowledge.

To find existing studies on a topic, researchers conduct a literature review. Reviewing research in an area helps scientists avoid duplicating others' work. It also allows them to find holes in existing knowledge and connect their research to broader theories in the field. Perhaps most importantly, reviewing previous research helps generate hypotheses that are based on patterns and findings already established in the field.

Imagine what it would feel like to design a study on personality and relationship status, gather the data, and then discover that dozens of similar studies already exist. Not only would you waste time and resources, but you would miss the opportunity to learn something new.

Near the top of the results, you should see a 2024 study by Julia Stern and her colleagues titled "Differences between lifelong singles and ever-partnered individuals in Big Five personality traits and life satisfaction." The research examined data from thousands of European participants over the age of 50 and found some interesting differences in the personality traits of people who are single and those who are married or in a long-term relationship. We encourage you to read the abstract to see whether your hypotheses align with their findings.

A useful feature of Google Scholar is the "Cited by" option that appears in the search results (see the red arrow in Figure 2.8). The "Cited by" feature shows papers that have referenced the study you are looking at. Reviewing these papers helps find related research because a study that is cited by other papers usually helped lay the groundwork for research on that topic.

You can also use Google Scholar for more advanced searches. By placing quotation marks around phrases like "Big Five personality" you can search for the exact term or phrase in your parentheses. In addition, you can add terms like "longitudinal" to find studies that track personality and relationship status over time. You can also restrict your search by year (e.g., since 2018) or by authors who do research on a topic.

Beyond Google Scholar

While Google Scholar provides access to a lot of literature, you may encounter paywalled content. When this happens, your school's library resources can be useful. Most academic libraries subscribe to many journals and databases. These databases provide students and faculty with access to articles that would otherwise be too expensive to purchase individually.

Commonly used institutional databases include PsycINFO (for psychology research), PubMed (for biomedical research), and Web of Science (for interdisciplinary research). These databases cover specific bodies of literature, and your university library likely provides access to these databases, along with tutorials on how to use them effectively.

If you ever come across a research paper that is not available on Google Scholar or through your university's electronic library, you can still get access to the article through what is called interlibrary loan. Your university's library likely has an interlibrary form available on its website. Simply fill it out by providing the article's information, and within a day or two they will e-mail you the article. Whether through Google Scholar, your institution's library, or interlibrary loan, you should have access to almost any academic article that you need for your research.

Connecting to Peer-Reviewed Research

As discussed in Chapter 1, scientific theories develop through systematic testing and refinement. In peer review, manuscripts submitted to academic journals are evaluated by other experts in the field (professional peers) who assess the quality and validity of the research before it is published.

When conducting a literature review, researchers typically prioritize peer-reviewed sources because they represent the most rigorously evaluated research. This does not mean peer-reviewed research is perfect. Scientific knowledge continues to evolve as new evidence emerges. But peer review does provide a foundation of methodologically sound studies on which to build.

By starting a research project with a thorough literature review, you connect your work to the broader scientific conversation about a topic. Your study becomes not just an isolated investigation, but a meaningful contribution to the collective understanding of human behavior.

The Pyramid of Knowledge

When doing a literature search, you are likely to come across multiple types of scientific publications, from research articles that focus on one specific question to whole books that provide an overview of an entire field. These different sources fit within the larger structure of scientific knowledge. This structure can be viewed as a pyramid, with each level building upon information from the levels below it (Figure 2.9).

At the base of the pyramid are individual peer-reviewed articles. These articles are studies that examined a specific research question. The study by Stern and colleagues (2024) comparing personality traits between lifelong singles and people in relationships is a good example. Each peer-reviewed article adds one block to the foundation of knowledge.

Moving to the second level, we find articles that synthesize research. These articles are often literature reviews and meta-analyses. These papers systematically gather and analyze findings from many individual studies. A literature review typically summarizes the findings from dozens of studies on a specific topic, whereas a meta-analysis uses statistical techniques to combine the results from dozens or even hundreds of studies. For instance, a meta-analysis by Malouff and colleagues (2014) analyzed data from multiple studies to examine the role that emotional intelligence plays in relationship satisfaction.

The third level of the pyramid contains books, edited volumes, and textbooks. These sources integrate information from review articles, meta-analyses, and individual studies and organize the knowledge into coherent frameworks. A textbook on personality psychology, for example, might include a chapter on how personality affects major life outcomes including relationships and happiness. While these sources may lack the details found in peer-reviewed journal articles, they provide a more complete understanding of a field.

Finally, at the apex of the pyramid are the major theoretical frameworks and paradigms that shape entire disciplines. In our example, this might include the Big Five theory of personality or broader theoretical perspectives on human relationships. These overarching theoretical frameworks are developed from the knowledge contained across all levels of the pyramid and can be found in every type of publication, including papers, overview articles, and books.

Together, the levels of the pyramid contain all the scientific knowledge about a topic.

After forming an idea and reviewing the literature, researchers design a study. This requires understanding the tools available to gather data. Generally, these tools fall into four categories: tools for gathering self-reported data, tools for measuring cognitive performance, tools for measuring physiological reactions, and tools that track behavior (Figure 2.10).

Survey Platforms: Gathering Self-Reported Data

At the heart of many behavioral studies are survey platforms like Qualtrics, SurveyMonkey, and Engage. These digital tools have largely replaced paper-and-pencil questionnaires and become the workhorses of modern behavioral research.

Survey platforms allow researchers to create professional-looking questionnaires without any programming knowledge. To conduct a study that examines whether there are personality differences between single and married people, we could easily use a platform like Qualtrics or Engage. We would select a measure of personality, like the Ten-Item Personality Inventory, and add questions about relationship status. Participants would then complete the survey from anywhere with an internet connection, on a computer or mobile device. With the data in hand, we could test our ideas.

What makes survey platforms so useful is their flexibility. They can present participants with various types of questions, ranging from multiple choice items ("I see myself as extraverted, enthusiastic" 1 = Strongly Disagree, 7 = Strongly Agree), to open-ended text responses ("In your own words, please describe what you think influences your relationship status?"). They also allow researchers to present pictures, video, and audio stimuli, and to control when these different materials are presented to participants. This flexible stimulus control allows researchers to create sophisticated studies that go beyond what people typically think of as a "survey." In Chapter 14, we provide information about how to create sophisticated research projects using survey platforms. Before that, virtually every chapter in this book demonstrates some basic techniques of working with Qualtrics and Engage.

Figure 2.11 shows what programming the TIPI questions looks like for a researcher in Qualtrics; Figure 2.12 shows how the questions appear to participants. It doesn't take much effort to present participants with professional looking surveys.

Beyond basic questions, survey platforms offer advanced features for executing sophisticated studies. The most important of these features is the ability to randomly assign participants to different conditions within a study and to randomize the order of stimuli. This makes it possible to conduct experimental research. For example, participants can be randomly assigned to watch different videos to assess whether they have a different emotional impact, and many other examples we will explore in Chapter 7.

Easy use and flexibility makes survey platforms extremely popular among behavioral scientists. When you encounter a research article reporting findings from a survey or experiment, chances are the researchers used one of these platforms to gather their data. In addition, some survey tools are beginning to integrate AI for even greater flexibility and sophistication. Chapter 8 examines one of these platforms and the doors it opens for research.

Performance Tools: Measuring Cognitive Processes

While survey platforms excel at gathering self-reported data about attitudes, beliefs, and experiences, behavioral scientists often need to measure aspects of human behavior that people cannot easily report. To do this, they use specialized software that measures cognitive performance (Figure 2.13). Software packages like E-Prime and SuperLab allow researchers to present stimuli to participants and measure their responses with millisecond precision.

For instance, a researcher who is interested in how quickly people can identify emotions might design an experiment where pictures of people with different facial expressions flash on a computer screen for just 200 milliseconds (1/5th of a second). Participants may be asked to press a key to indicate whether the face showed happiness, anger, fear, or surprise. The researcher wants to know how quickly people process the information and whether they are accurate. Can you imagine the researcher asking people to self-report how fast they recognized the emotions? Most people would have no clue. For that reason, behavioral scientists don't ask. They measure.

Performance measures are commonly used in fields like cognitive psychology and cognitive neuroscience to study mental processes such as attention, memory, and decision-making. Although these tools require more technical knowledge than survey platforms, they are necessary for many research questions.

For example, we might use these tools to explore whether single and partnered people differ in how quickly they recognize emotional expressions—an aspect of social cognition that could both influence and be influenced by experiences in relationships. Such a study could provide a more complete picture of the psychological differences associated with relationship status.

Physiological Measurement

Beyond what people say or how they perform on cognitive tasks, their bodies provide a wealth of information about what they are thinking, feeling, and doing. Behavioral scientists refer to measures of bodily response as "physiological measurements," and these measures provide access to things happening in the body that are often outside of conscious awareness (Figure 2.14).

Physiological research relies upon equipment such as wristbands, heart-rate monitors, eye-trackers, skin conductance devices, and software (e.g., BioPac, Empatica) to measure and interpret people's reactions to different situations. To see the value of these measures, imagine a researcher studying stress in social situations. While participants might report feeling "a little nervous" during a job interview, their bodies may tell a more detailed story. Measuring physiological responses, the researcher might see that the person's heart rate, blood pressure, and skin conductance (how much the skin sweats) are equivalent to levels seen during light physical exercise, such as walking briskly up a flight of stairs. Even though the person's biological responsiveness may show significant signs of stress, they can be subjectively interpreted as "a little nervous."

In fields like biological psychology and neuroscience, physiological measurement tools are critical for understanding how people's emotional and physiological states interact to influence thought and behavior. In a study on personality and relationships, for instance, we might use physiological measures to examine whether singles and people with a partner differ in physiological reactions when viewing images of couples or thinking about past relationships. Perhaps people with certain personality profiles show a distinct physiological reaction when processing social or emotional information—patterns that might influence their relationship experiences.

While these tools provide valuable data the equipment can be expensive and requires expertise to use properly. Learning to use these tools typically requires a specialized course in biological psychology research methods and experience working in a research lab. These tools are a key aspect of behavioral research and can often complement the data gathered through surveys and performance measures.

Behavioral Measurement: Capturing Real-World Interactions

The final category of tools track how people behave in their everyday lives. Behavioral measurement tools allow researchers to observe and record real-world behaviors as they occur. One example comes from researchers studying global migration patterns, who used Meta to analyze anonymized data from billions of Facebook users worldwide (Figure 2.15). By tracking changes in users' location data over time, they created detailed maps of migration flows between countries.

Measuring behavior extends beyond digital tracking. Economists analyze purchasing records to study consumer behavior. Urban planners use traffic cameras to study pedestrian movement. Environmental psychologists place sensors in homes to measure energy usage, as you will see in Chapter 3. In each case, researchers directly measure behavior rather than asking people to report what they did.

Similarly, social media sites, apps, and wearable devices can track people's behaviors across a variety of contexts and provide behavioral scientists with all kinds of data they could never collect in a laboratory. During the early stages of the COVID-19 pandemic, for instance, researchers tracked searches for symptoms like "loss of smell" and "loss of taste" to identify potential outbreaks before official numbers were reported (Cherry et al., 2020). In many cases, behavioral data can be compared to people's self-reported data to see if the two align. If someone says in a survey that they "occasionally check their phone throughout the day" but an app reveals that they actually check 150 times in an average day, the person is either unwilling or unable to accurately report on their behavior.

Making sense of so much data often requires advanced training in statistical analysis. This book will help you develop a foundation you can build on later. When thoughtfully implemented, behavioral measurement tools complement other research approaches by providing objective data about what people actually do.

After the data for a study are collected, the next step is always to analyze it with statistical tests. In other words, statistical analysis is universal in research.

Traditional Data Analysis Tools

In Chapter 1, you saw the basics of data analysis when you entered your TIPI scores into Google Sheets. The spreadsheet calculated simple statistics like the average score for each personality trait. While spreadsheet programs like Excel and Google Sheets work for basic calculations, most behavioral research requires more advanced statistical tools. These tools rely on specialized software capable of performing complex calculations on large datasets.

One of the most widely used programs in the behavioral sciences is SPSS (Statistical Package for the Social Sciences). SPSS is particularly popular in introductory statistics and research methods courses because of its intuitive user interface (Figure 2.16). In addition to the rows and columns of data you can find in any spreadsheet, SPSS has menus and icons across the top that make it easy to perform all kinds of data operations and analyses. Throughout this book, we use SPSS to analyze data. You will get your first crack at SPSS in Chapter 3.

Another powerful tool is R, a free and open-source program that offers tremendous flexibility for statistical analyses. Simpler alternatives like JASP and Jamovi are often used to teach students the basics of data analysis. They combine an intuitive user interface with powerful statistical capabilities that include built-in explanations of statistical tests, making them excellent for beginners.

Regardless of which statistical package a researcher chooses, the goal remains the same: find meaningful patterns in the data that answer the research question. If we were examining a dataset to look for differences in Big Five traits across relationship status, we might first create an average score for each participant across the five personality traits. Then, we would divide participants into either a singles group or a partnered group (anyone who had ever committed to a lifelong partner). After preparing the file, we would perform a statistical test to determine whether the average differences between the two groups were meaningful. If we found "statistically significant" results, we would share them with other researchers, as Stern and colleagues did in their paper (2024). They found that in addition to having lower levels of life satisfaction, lifelong singles were less extraverted, less conscientious, and less open to experience compared to partnered people. How do their results align with your hypotheses?

Statistical packages are not only useful for making sense of data, they also help researchers create visual ways to communicate the results. Charts, graphs, tables, and figures are an important way to effectively communicate the main findings of a research study. In later chapters, we will learn how to use these data analysis tools to test specific hypotheses, create visualizations, and draw conclusions from data.

AI-Powered Tools for Data Analysis

In recent years, artificial intelligence has transformed many aspects of research, including data analysis. Modern AI tools offer specialized features that can perform sophisticated statistical analyses. One notable example is ChatGPT's Data Analyst feature, available in the paid version of the platform.

What makes AI-powered analysis tools particularly appealing is their ability to understand commands in everyday language. For example, you might upload a dataset and type, "Calculate the mean scores for each personality trait, create a bar chart comparing them, and conduct an ANOVA to see if the groups are different from each other." Then, like magic, the AI will execute the commands.

If you have access to the paid version of ChatGPT, you can download the Google Sheet from Chapter 1 as an Excel file, upload it to the Data Analyst, and ask it to calculate averages for each personality trait. Then you can ask it to create a bar chart. The result should match what you saw in the original spreadsheet, but you would achieve it through conversational commands.

The Data Analyst feature can handle all the analyses we will cover in this book. We have verified that its output matches SPSS results across all our example projects, from basic descriptive statistics to more complex inferential statistics.

There are several advantages to using AI for data analysis. The conversational interface reduces the learning curve typically associated with statistical software. You can iterate quickly, refining analyses based on initial results and, most importantly, the AI can also explain statistical concepts and answer your questions as it works!

However, as with all AI applications, there are important considerations. AI-powered analysis tools should be used in collaboration with knowledgeable professionals, such as your course instructor. Traditional tools like SPSS also provide more transparency that makes it easier to verify the accuracy of results, particularly for advanced analyses. The AI might also select analytical approaches that seem reasonable but don't align with best practices for your specific research question.

Throughout this book, we focus primarily on SPSS while acknowledging that AI alternatives exist. We encourage those with access to Data Analyst or similar tools to explore them as complementary resources, especially when getting started with a new analysis or seeking to understand statistical concepts in plain language.

The work of modern behavioral research doesn't end once the data are analyzed. In fact, it is only then that the work of clearly communicating the findings begins.

Behavioral scientists often communicate about research in public lectures, conference presentations, and journal articles. Common tools like PowerPoint and word processors help researchers organize and disseminate their findings. Chapter 16 provides some advice about writing journal articles.

In addition to publishing research findings, researchers often use platforms like the Open Science Framework (OSF) to organize, store, and share their data and materials (Figure 2.17). The OSF represents an important shift in how behavioral science operates. In the past, researchers typically shared only their results through published articles. The raw data, survey questions, analysis code, and other materials that went into the research remained private. But this practice created problems.

The most important problem was that other scientists often could not easily verify published findings. This became known as the replication crisis. The open science approach helped address this problem by providing an online platform where researchers can share their work. Scientists upload their materials, data, and analysis scripts to OSF. These materials become available to anyone interested in understanding or extending the research. This openness promotes scientific transparency and enables other researchers to replicate findings.

Throughout this book, you will access research materials from the Research in the Cloud OSF page. You will find survey files you can import directly into Qualtrics and Engage, datasets you can analyze, and scripts showing exactly how we conducted analyses. Using OSF will give you hands-on experience with another one of the online tools that define modern behavioral research. When you work with materials on OSF, you are participating in the culture of open science.

Finally, beyond the OSF, behavioral scientists use tools that help them submit research protocols for ethical review; track references and citations of a manuscript; submit grant applications and journal articles for review; and communicate with other researchers while collaborating on a project. None of these tools are essential, but they support the research process and make tasks more efficient.

Summary

In this chapter, you have taken the next steps toward becoming a behavioral researcher. You experienced what online research is like by participating in studies on Connect. Then, you created a researcher account on Connect, taking you closer to launching your own projects. And, finally, you walked through the steps of the research process while learning about the tools researchers use in their work.

Collectively, the integration of so many advanced tools has changed how behavioral research is conducted. A single study might use recruitment platforms like Connect to find participants, survey tools like Qualtrics or Engage to create the study, statistical software like SPSS to analyze the data, and visualization tools to present the findings. Then, the research team might use a data repository like OSF to make their materials and data available to others while issuing a pre-publication version of their manuscript while they await feedback from the traditional peer-review process. Altogether, this technological ecosystem enables researchers to address more complex questions about human behavior and to do so with greater speed and effectiveness than ever before.

But it's important to remember that these tools, while powerful, are just that: tools. Their value, like that of a hammer, tractor, scalpel, or fermenter, comes from the ability of a knowledgeable professional putting them to use toward a thoughtfully chosen goal. Even the most sophisticated technology cannot compensate for poor research design or flawed methodology.

Thus, while you will have many opportunities to practice with the tools described above throughout the rest of this book, it is important that you also learn the fundamentals of good research design. Understanding how to build a descriptive, correlational, or experimental research project to answer a question you are curious about is a skill that will serve you well. That is why each of the upcoming chapters contains many opportunities for you to develop your knowledge and skill as a behavioral researcher.