Generation Me--Revised and Updated

by Jean M. Twenge

  Appendix

  * * *

  Further Details about the Scientific Studies in This Book

  If you’ve turned here, you’re probably curious about the psychological terms and search methods I’ve used. Here I’ll address a few topics that often raise questions. If you’d like the real nitty-gritty on the statistics and methods of the studies, I recommend finding my (and others’) scientific-journal articles through Google Scholar, the Web of Science, PsycInfo, or another database. Many of these are listed in the notes section.

  Personality traits and attitudes. What exactly are traits? Basically, they are the internal attributes of people that cause their behavior. Traits summarize the way someone behaves across most situations. For example, someone who expresses her opinions and speaks up when she feels wronged would be high in the trait of assertiveness. Most questionnaires ask fairly specific questions; one measure of assertiveness asks, “If a friend unfairly criticizes you, do you express your resentment there and then?”

  So if a questionnaire wanted to measure the trait of “silliness,” it might ask about a number of different silly behaviors, such as making funny faces, asserting amusingly absurd statements, and playacting unconsciousness. We learn to group behaviors this way early. If you do all of the things I just listed, for example, you will eventually hear from your child, “Mommy, you’re silly!” So when you read that a trait (such as anxiety or assertiveness) has changed, remember that this seemingly abstract concept is based on these more specific questions about actions in real life. Traits are about behavior.

  I also mention attitudes, such as attitudes toward working women, support for same-sex marriage, or confidence in large institutions. In the psychological sense, an attitude is a favorable or unfavorable opinion of a group or a group’s actions. If you have a strong attitude and express it, you have “attitude” in the more colloquial sense. Here, though, I will focus mostly on attitudes toward the roles of certain groups. For example, attitudes toward the roles of women have changed quite a bit. Like the questionnaires on traits, attitude questionnaires also ask specific questions (e.g., Do you agree or disagree that “Sons in a family should be given more encouragement to go to college than daughters”?). Attitudes are about beliefs, often deeply held beliefs.

  The content of the questionnaires used in the studies. As you’ve seen, some of the generational data comes from people’s responses to questionnaires. These questionnaires usually consist of a number of statements that people are asked to agree or disagree with. Sometimes they can answer only yes or no, and other times they give their response on a scale, say from 1 to 5. A few questionnaires (such as the Narcissistic Personality Inventory) use what’s called forced choice, meaning you have to choose which statement out of two you agree with the most. The statements tend to ask about specific behaviors, feelings, or opinions. It’s fairly difficult to know what these specifics are, though, if you just hear the name of a scale. To give you a better idea of what the questionnaires are actually measuring, sometimes I quote at least a few items from the questionnaires I studied when I describe the results. It helps, for example, to know that one of the items on a scale measuring narcissism is “If I ruled the world, it would be a better place.” This shows that the questionnaire isn’t just talking about physical vanity. When my students fill out the narcissism questionnaire in class, someone will inevitably raise his hand right away, announce that he scored high on narcissism, say, “I think that . . . ,” and go on and on about himself and his ideas. It increases my confidence in personality questionnaires every time.

  Changes in college populations. Especially in this updated edition, many of the studies are based on high school samples. But other studies are based on college students. But haven’t those samples changed a lot recently? Surprisingly, the answer is no. A large yearly study found that the median income of college students’ parents, when adjusted for inflation, did not change between 1965 and 2013. College students come from middle-class and upper-middle-class families, and this has not changed much over time. There has not been a large influx of working-class students, partially because yearly tuition increases have kept four-year colleges out of reach for many. Changes in racial composition are also small, with minority-student enrollment increasing only a few percentage points in the last forty years. College enrollment for black students actually declined during the 1980s. Most college-student samples remain mostly white, just as they were in earlier eras. However, large changes have occurred in gender composition: 36% of college students were women in 1958, compared to 58% in 2012. To correct for this, I analyzed the data within gender; for example, I looked at the changes in women’s anxiety scores over time, and men’s anxiety scores over time. In most cases, the results were similar. College samples do have more Latinos and Asian Americans than they once did. Interestingly, these two groups tend to score lower on positive self-views, so if changes in ethnic composition caused the differences, self-esteem and narcissism would have gone down over the generations instead of up.

  In many of these studies, I also looked at samples of schoolchildren. Children of elementary school and middle school age are almost always enrolled in school, and their samples have not changed much in composition (compared to the small changes in college populations). The new studies also draw from a large sample of high school students. These samples are also more diverse and thus better represent the country as a whole, and the high school sample is nationally representative, meaning it exactly reflects the composition of high school students across the country. When I compared the results for children and high school and college students, they usually showed exactly the same pattern. This gave me further confidence that the changes were not an artifact of shifts in college-student samples. It also showed that children were not immune to these larger social changes—the generational shifts appeared in samples of children as young as 9. Kids are absorbing the ways of their society, and they are doing it at an early age.

  Willingness to admit to problems. Perhaps the changes in the questionnaires happen because people now have fewer qualms admitting to problems. However, this is unlikely to account for much of the change. First, I collected data on a scale measuring socially desirable responding—how much people change their questionnaire answers in order to look good in the eyes of others. When I matched that to data on other questionnaires, they weren’t correlated over time. So changes in traits happened independent of people’s comfort with admitting to things. Also, the questionnaires across all of these studies were given on paper and not in interviews, and they’re anonymous—respondents don’t put their names on them. The questionnaires also ask about specific symptoms (“Some unimportant thought runs through my mind and bothers me”) rather than asking point-blank something like “Are you anxious and depressed?” The responses to all of the symptoms are added up to form the score, so the respondent only admits to small parts of a problem at a time. Finally, the changes described in this book are diverse. Some of them are in “good” traits (such as self-esteem), but others are in “bad” traits (such as anxiety), and some (such as who controls your fate, or narcissism) have questions worded with no obvious “good” or “bad” answer. If people were more comfortable admitting to bad things, we’d expect to see change only in traits that are considered undesirable, but the changes show up in all kinds of characteristics.

  The ecological fallacy. Some of the studies in this book use a method I call cross-temporal meta-analysis (CTMA), which gathers the average scores of samples over time. But, some have asked, doesn’t this overestimate the change, as means differ less from each other than individual scores? (This is sometimes called the ecological fallacy, or alerting correlations.) It does not—starting with the first CTMA, I computed the effect size using the individual standard deviation (SD), not the SD of the means. Thus the effect sizes of these studies are not overestimates—they rely on the same measure of variance that studies with individual data do. Some critics still try to clai
m that CTMAs commit the ecological fallacy, but they do not, and have not from the very first one published in 1997. It is puzzling that some still make this patently false claim.

  The effects of specific events (such as 9/11/01 and the Great Recession of the late 2000s). In many studies, I have to estimate the year data were collected based on the date of publication, since most researchers don’t report the exact year in their articles. Statistically speaking, this means, if anything, that changes are stronger than I’ve found (as the imprecise year introduces more uncertainty). Even if I could pinpoint the exact year, though, I’m guessing that the change would still be slow and steady. Specific events don’t have as much influence as the social climate they create over a period of years. Social change does not happen overnight, particularly in traits and attitudes—characteristics of people that tend to be fairly stable. In fact, I’ve found in several studies that the change in traits lags about ten years behind statistics such as divorce rate, age at marriage, or crime rate. This suggests that we absorb the environment that surrounds us when we are about 10 years old. Morris Massey, who traveled the country lecturing about generational differences in the 1970s and ’80s, calls this “Where were you when you were 10?” and my data support this idea. The environment we experience as children stays with us our whole lives.

  In other cases, it was possible to see how multiyear changes such as the recession of 2007–9 impacted young people. My coauthors and I have published one paper on that topic and are working on several others. The recession provided a cultural reality check that has increased young people’s concern for others and caring about social issues. In the coming years, it will be fascinating to see if it has also reduced narcissism—and, if so, whether narcissism will come back once the economy does.

  Age differences versus generation differences. You might also wonder why all of the library work and huge database trolling was necessary. Why not give these personality questionnaires right now to a big sample of people of different ages? Shouldn’t that tell us how generations differ? If we did that, though, we wouldn’t know if age or generation was causing the difference. First, I’ll show you some examples of differences that are clearly one or the other. Let’s say we find that people in their 20s have more energy than people in their 40s. Does this mean that the generation of people in their 20s is more energetic than those in their 40s and always will be?

  Probably not; every generation finds that their energy level decreases as they age. This is clearly a difference due to age and not to generation. In contrast, consider your skill with Internet message boards, ATM machines, and programming your DVR. If you’re a little older, these might not be your strong suit. (My parents, for example, still don’t use ATMs—they go to the bank. I just don’t get it.) This is clearly a difference based on generation and not age, because people who are young now will not suddenly forget their current technology skills when they are older.

  But in a one-time sample of people of different ages (called a cross-sectional study), it’s impossible to tell if differences are due to age or generation. If a study finds, for example, that people in their 20s are more anxious than people in their 40s, this could be due to aging (maybe people become less anxious as they get older) or to generation (maybe the younger generation will always be more anxious than the older generation). The method I use gets around this problem by looking at samples of people of the same age collected at different points in time. Each of these data sources is a snapshot in time of what each generation was like when young, either children or young adults in college. Because age is held constant, generation and time are clearly the forces at work. Of course, it’s possible that people of all ages showed these same changes; if so, that would be a time-period effect rather than a generational effect, and I can’t rule that out. But anyone who has observed people knows that young people are more susceptible to change. For example, older people don’t use computers as proficiently because they didn’t learn how when they were young. Social change hits the young first, and I wanted to see how that happened from the 1950s to the present.

  Without a time machine, I couldn’t get these data myself. Since so many psychological scales have been used over and over, though, the library became my time machine. Those old journals and dissertations were my looking glass to the past, showing me how things used to be and how they are now. Since the last edition, I’ve also trolled through the enormous datafiles of over-time surveys such as the Monitoring the Future, the American Freshman, and the General Social. Amid screen after screen of numbers, the picture of social change slowly emerged.

  Acknowledgments

  * * *

  As a child, I loved to read so much that I would often stay up past my bedtime to finish a book. My practical Midwestern parents, however, had the radical idea that I should get enough sleep. I tried to finish a book in the bathroom once but was promptly busted. After that, I learned to leave on the aquarium light, which was just bright enough to read by but not bright enough to shine under the door and alert the parental units to my stolen reading time. I would lie at the foot of my bed, late into the night, turning the magic pages in the dim light.

  Twenty-five years later, I have read the acknowledgments sections of so many books that I can recite the typical one from memory (“It is fiction that it takes only one person to write a book . . . to my agent extraordinaire . . . to my 546 closest friends [Aaron, Adam, Alice, Amanda . . . (insert 540 names here) . . . Zachary, and Zelda] . . . and finally, to my spouse for cooking all the meals for the last year/raising several children by himself while I slaved away at this godforsaken project/reading the whole book/typing the whole book/writing the whole book”). After writing a book myself, I have suddenly realized why acknowledgment sections all sound the same: you’re a massive pain in the ass when you’re working on a book, and a little ink is small consolation for all of the people who had to put up with it. I was no different, so my acknowledgments will probably sound no different.

  I was a pain partially because writing this book was the most fun I’ve had at work in years, and enthusiasm, although infectious, can get annoying. Everyone, please forgive me. It’s just that any project in which I can quote Us Weekly and academic research on the same page is a damn good time.

  I owe my first, large sums of gratitude to those who made writing this book even more fun than it was already. Brooke Wells wrote her amazingly thorough master’s thesis on changes in sexuality over time, providing the data for chapter 6 and enabling me to spend several enjoyable weeks writing about sex. How can I not be grateful for that? Brittany Gentile and Joshua Foster provided much-needed updates to data, and Kristin Donnelly, Elise Freeman, Joshua Grubbs, Julie Exline, and Nathan Carter put in crucial work on new projects. W. Keith Campbell, my colleague, coauthor, and friend, has sharpened my thinking on just about everything for fifteen years now, including several parts of this book. I’m not sure I would have made it through the jungle of academia without Keith’s hatchet of wry irony to clear the way. Next conference, let’s skip the symposia and go talk and eat doughnuts again. I learned more that way anyway.

  I am also profoundly grateful to the hundreds of young people who opened their lives to me by contributing stories for the book. A large thanks in particular to those who volunteered their time by submitting material through the www.generationme.org and www.jeantwenge.com websites. Your honesty, insight, and eloquence made this book immeasurably stronger. Your stories and opinions brought your generation to life.

  My agent, Jill Kneerim, is the reason I got to write this book in the first place. I am amazed, and truly grateful, at how she was able to guide me so encouragingly toward a book proposal that bore no resemblance to my first pathetic attempts. I can only hope to someday develop even an ounce of Jill’s ability to pleasantly guide someone into doing her best. My editor, Leslie Meredith, has been equally delightful. Writing your first book can be terrifying, but all of that melted away once I saw Leslie’s enthusiasm for m
y first few chapters. ;-) Long live the deadline emoticon! 8-] The brilliant Brettne Bloom came up with the title and thus deserves a huge high five.

  I have been blessed with truly outstanding academic mentors: Roy Baumeister, Jennifer Crocker, Susan Nolen-Hoeksema, Randy Larsen, Abby Stewart, Dianne Tice, and David Winter all helped guide some aspect of this project. Susan showed me the research on changes in depression and generally gave fantastic advice. Her passing in 2012, much too early, saddened all of us in the field, but especially those who had the pleasure of knowing her personally. David guided me through the uncharted territory of lagged analyses and obscure personality measures, gently shaping my overenthusiastic grad-student ideas into a more conventional psychology dissertation. We also had quite a good time just chewing the fat. Roy was the best postdoctoral adviser I could have asked for, and a good friend. His pioneering research on the downsides of self-esteem also strengthened the book considerably. And in the beginning, when I was only 15, Melissa McMillan-Cunningham and Patrick McCann both told me I would someday write a book. We all thought it would be a book of poetry, but I hope this is just as good.

  Charles Im and Liqing Zhang spent hours upon hours in the library collecting some of the data I use here, and they deserve ample praise and gratitude. Your hard work made these projects possible. Mark Reid photocopied numerous articles on changes in depression and provided inspiration through his excellent writing. A true man of many talents, he also designed the www.generationme.org website, where I collected stories from young people around the nation. Andrew Chapman designed my newer websites (jeantwenge.com and igenconsulting.com) and is a Web-design genius if you ask me.