25

Theories give us lenses for interpreting behavior. Research allows us to find the scientific truth. I already touched on the research technique designed to determine the genetic contributions to behavior. Now let’s sketch out the general research strategies that developmental scientists use.

What impact does poverty have on relationships, personality, or physical health? What forces cause children to model certain people? Does a particular intervention to help improve self-efficacy really work? To answer any question about the impact one condition or entity (called a variable) has on another, developmentalists use two basic research designs: correlational studies and true experiments.

In a correlational study, researchers chart the relationships between the dimensions they are interested in exploring as they naturally occur. Let’s say you want to test the hypothesis that parents who behave more lovingly have first graders with superior social skills. Your game plan is simple: Select a group of children by going to a class. Relate their interpersonal skills to the nurturing that parents provide.

Immediately, you will be faced with decisions related to choosing your participants. Are you going to explore the practices of mothers and fathers or mothers alone, confine yourself to a middle-class group, consider two- versus one-parent families, look at a mix of ethnicities or not? You would need to get permission from the school system. You would need to get the parents to volunteer. Are you choosing a representative sample—a group that reflects the characteristics of the population about whom you want to generalize?

Then you would face your most important challenge—accurately measuring your variables. Just as a broken thermometer can’t tell us if we have a fever, if we don’t have adequate indices of the concepts we are measuring, we can’t conclude anything at all.

With regard to the parent dimension, one possibility might be to visit parents and children and observe how they relate. This technique, called naturalistic observation, is appealing because you are seeing the behavior as it occurs in “nature,” or real life. However, this approach presents a huge practical challenge: the need to travel to each home to observe each family on many occasions. Plus, when we watch parent–child interactions, or any socially desirable activity, people try to act their best. Wouldn’t you make an effort to act especially loving if a psychologist arrived at your house?

The most cost-effective strategy would be to give the parents a questionnaire with items such as: “Do you make an effort to kiss, hug, and praise your daughter? Is it important to avoid yelling at your child?” This self-report strategy, in which people evaluate their behavior anonymously, is the main approach researchers use with adults. Still, it has its own biases. Do you think that people can report accurately on their activities? Is there a natural human tendency to magnify our positive behaviors and minimize our negative ones?

Now, turning to the child side of your question, one reasonable way to assess social skills would be to have teachers evaluate each student via a questionnaire: “Does this child make friends easily?”; “Does he relate to his peers in a mature way?” Or, we could ask children to rank their classmates by showing photos: “Does Calista or Cory get a smiley face?”; “Pick your three best friends.” Evaluations from expert observers, such as teachers, and even peers, are often used to assess concepts such as popularity and personality during the childhood years.

26

Table 1.5 spells out the uses, and the pluses and minuses, of these frequently used ways of measuring concepts: naturalistic observation, self-reports, and observer evaluations. Now, returning to our study, suppose you found a relationship, that is, a correlation, between having nurturing parents and children’s interpersonal skills. Could you infer that a loving home environment causes children to socially flower? The answer is no!

To rule out these confounding forces, the solution is to conduct a true experiment (see Figure 1.4). Researchers isolate their variable of interest by manipulating that condition (called the independent variable), and then randomly assign people to either receive that treatment or another, control intervention. If we randomly assign people to different groups (say, like tossing a coin), there can’t be any preexisting differences between our participants that would bias our results. If the group does differ in the way we predict, we have to say that our intervention caused the particular result.

The problem is that we could never assign children to different kinds of parents! If, as Figure 1.4 suggests, developmentalists trained one group of mothers to relate in more caring ways and withheld this “intervention” from another group, the researchers would run into ethical problems. Would it be fair to deprive the control group of that treatment? In the name of science, can we take the risk of doing people harm? Experiments are ideal for determining what causes behavior. But to tackle the most compelling questions about human development, we have to conduct correlational research—and control as best we can for competing explanations that might bias our results.

27

Experiments and correlational studies are standard, all-purpose research strategies. In studying development, however, we have a special interest: How do people change with age? To answer this question, scientists also use two research designs—cross-sectional and longitudinal studies.

Cross-Sectional Studies: Getting a One-Shot Snapshot of Groups

Because cross-sectional research is relatively easy to carry out, developmentalists typically use this strategy to explore changes over long periods of life (Hertzog, 1996). In a cross-sectional study, researchers compare different age groups at the same time on the trait or characteristic they are interested in, be it political attitudes, personality, or physical health. Consider a study that (among other questions) explored this interesting issue: “How do our feelings about human nature change with age?”

Researchers gave 2,138 U.S. adults a questionnaire measuring their beliefs in a benevolent world (Poulin & Silver, 2008). Presented with items such as “Human nature is basically good,” people ranging in age from 18 to 101 ranked each statement on a scale from “agree strongly” to “disagree.” As you scan the findings in Figure 1.5, notice that the youngest age group has the most negative perceptions about humanity. The elderly feel most optimistic about people and the world. If you are in your early twenties, does this mean you can expect to grow less cynical as you age?

28

Not necessarily. Perhaps your cohort has special reasons to feel suspicious about human motivations. After all, today the media delights in exposing the cheating and lying of authority figures, from senators to school principals. Previous cohorts of young people were never exposed to this drumbeat of messages highlighting human nature at its worst. In fact, if we conducted this same poll during the 1950s (in the Eisenhower era of Leave It to Beaver and Father Knows Best) we might find the opposite pattern: Positive feelings about human nature were highest among the young and declined with age!

The bottom line is that cross-sectional studies give us a current snapshot of differences among cohorts (or age groups); but they don’t necessarily tell us about real changes that occur as we grow old.

Cross-sectional studies have a more basic problem. Because they measure only group differences, they can’t reveal anything about the individual differences that give spice to life. If you are a real pessimist, will your worldview stay the same as you age? What influences might make people feel better about humanity during adulthood, and what experiences might make people feel worse? To answer these questions about how individuals develop, as well as to look at what makes for specific changes, it’s best to be on the scene to measure what is going on. This means doing longitudinal research.

Longitudinal Studies: The Gold-Standard Developmental Science Research Design

In longitudinal studies, researchers typically select a group of a particular age and periodically test those people over years (the relevant word here is long). Consider the Dunedin Multidisciplinary Health and Development Study: An international team of researchers descended on Dunedin, a city in New Zealand, to follow more than 1,000 children born between April 1972 and March 1973, examining them at two-year intervals during childhood and roughly every three years after that. At each evaluation, they examined participants’ personalities and looked at parenting practices and life events. The scientists are now tracking these babies as they move into middle age (Dunedin Multidisciplinary Health and Development Research Unit, 2014).

The outcome has been an incredible array of findings related to personality and psychological problems. Can we predict adult emotional difficulties as early as age 3? Do chronic anxiety and depression produce cellular damage as we travel through adolescence and early adult life (Shalev and others, 2014)? As you will learn in Chapter 12, scientists have discovered that a personality trait called conscientiousness, measured during emerging adulthood, powerfully predicts following good health habits—from not smoking, to exercising, and to watching one’s diet as participants travel into middle age (Israel and others, 2014).

Moreover, because they are using cutting-edge DNA technology, this landmark research was the first to identify candidate genes that may make us more or less responsive to wider world stress. Plus, like other longitudinal research, the Dunedin Study offers a crystal ball into those questions at the heart of our field: How will I change as I get older? When should we worry about children, and when should we not be concerned?

Longitudinal studies are exciting, but they have their own problems. They involve a tremendous amount of time, effort, and expense. Imagine the resources involved in planning this particular study. Think of the hassles involved in searching out the participants and getting them to return again and again to take the tests. The researchers must fly the overseas Dunedin volunteers back for each evaluation. They need to reimburse people for their time and lost wages. These logistical and financial problems become more serious the longer a study continues. For this reason, we have hundreds of studies covering infancy, childhood, or defined segments of adult life, such as the old-old years. Only a precious handful trace development from childhood to later life.

29

The difficulty with getting people to return for testing leads to an important bias in itself. People who stay in longitudinal studies, particularly during adulthood, tend to be highly motivated. Think of which classmates are going to attend your high school reunion. Aren’t they apt to be the people who are successful, versus those who have made a mess of their lives? Participants in longitudinal research are typically elite, better than average groups. While they offer us unparalleled information, these gold-standard studies have their biases, too.

So to summarize, when you are scanning the findings in our field, keep these concerns in mind:

Developmental scientists are attuned to these issues. In conducting correlational studies, they typically try to control for other influences that might explain their findings. They are apt to use several measures, such as teacher ratings, formal questionnaires, and peer input, as well as direct observations, to make sure they are measuring their concepts accurately. As you will see throughout this book, current developmental research has an international flavor, with researchers from nations as different as Iran and Ireland or China and Cameroon offering country-by-country insights on core topics in our field. Still, in addition to becoming more global, our research is getting up close and personal, too.

Quantitative research techniques—the strategies I have been describing, using groups of people and statistical tests—are the main approaches that researchers use to study human behavior. In order to make general predictions about people, we need to examine the behavior of different individuals. We need to pin down our concepts by using scales or ratings with numerical values that can be tallied and compared. Developmentalists who conduct qualitative research are not interested in making numerical comparisons. They want to understand the unique lives of people by conducting in-depth interviews. In this book, I will be focusing mainly on quantitative research because that is how we find out the scientific “truth.” But I also will highlight the increasing number of qualitative interview studies to put a human face on our developing life.

30

This discussion brings me back to the letter on page 2 and my promise to let you in on my other agendas in writing this text. Because I want to teach you to critically evaluate research, in the following pages I’ll be analyzing individual studies and—in the How Do We Know features that appear in some chapters—focusing on research-related issues in more depth. To bring home the personal experience of the lifespan, I’ve filled the chapters with quotations, and—in the Experiencing the Lifespan boxes—occasionally interviewed people myself. To bring home the principle that our lifespan is a work in progress, I’ll be starting many chapters by setting the historical and cultural context. To emphasize the power of research to improve lives, I’ll conclude most sections by spelling out interventions that improve the quality of life.

This book is designed to be read like a story with each chapter building on concepts and terms mentioned in the previous ones. It’s planned to emphasize how our insights about earlier life stages relate to older ages. I will be discussing three major aspects of development—physical development, cognitive development, and personality and social relationships (socioemotional development)—separately. However, I’ll be continually stressing how these aspects of development connect. After all, we are not just bodies, minds, and personalities, but whole human beings!

While I want you to share my excitement in the research, please don’t read this book as “the final word.” Science—like the lifespan—is always evolving. Moreover, with any research finding, take the phrase “it’s all statistical” to heart. Yes, developmentalists are passionate to make general predictions about life; but, because human beings are incredibly complex, at bottom, each person’s lifespan journey is a beautiful surprise.

Now, beginning with prenatal development and infancy (Chapters 2, 3, and 4); then moving on to childhood (Chapters 5, 6, and 7); adolescence (Chapters 8 and 9); early and middle adulthood (Chapters 10, 11, and 12); later life (Chapters 13 and 14); and, finally, that last milestone, death (Chapter 15), welcome to the lifespan and to the rest of this book!

Question 1.11

1. Who is conducting a cross-sectional study?

2. Who is using naturalistic observation?

3. Who is conducting a correlational study?

4. Who can prove that eating a lot of sugar causes problems—but is doing an unethical study?

5. Who is going to have a huge problem with dropouts?

6. Who can tell you that if you are a sugar junkie in your twenties, you might still be eating an incredible amount of sugar (compared to everyone else) as you age?

a. David; b. Alicia; c. Alicia; d. Betty; e. Calista; f. David

Question 1.12

After coming up with your hypothesis, you would need to adequately measure your concepts—choosing the appropriate tests. Your next step would be to solicit a large representative sample of a particular age group, give them these measures, and retest these people at regular intervals over an extended period of time. In addition to the investment of time and money, it would be hard to keep track of your sample and entice participants to undergo subsequent evaluations. Because the most motivated fraction of your original group will probably continue, your results will tend to reflect how the “best people” behave and change over time (not the typical person).