share
interactive transcript
request transcript/captions
live captions
download
|
MyPlaylist
[APPLAUSE] RAND CONGER: Well, thank you, John, for that introduction. And I want to thank the department, faculty, and students for inviting me to be part of this lecture series. I never had the opportunity to meet Dr. Ricciuti, but I know some of his work. And I'm very impressed and very flattered that I would be invited to give a talk at a series that is sponsored by somebody who was so beloved by the department.
I haven't met anybody here yet who's had anything but praise and admiration for Dr. Ricciuti. So I'm very, very pleased to be here to be part of this series, and very pleased to be invited along with the other folks who've come and given talks. It's a very good, very outstanding group of scholars who you've invited and who've come to talk about their work.
So today, some of you may be familiar with the work that I've done on family stress processes and how family stress processes influences parenting and child outcomes. Today I'm going to go in a little bit different direction. I'm actually going to talk about how advantages, social and economic advantages, actually promote competent development, particularly in terms of the development of conscientiousness.
Now, I don't cite a lot of papers. Actually, I don't know if I cite any papers in the PowerPoint slides. But a good place for people to start if they're interested in the issue of conscientiousness is a June, 2014 special section in Developmental Psychology where there's a series of wonderful papers in that special section on what we know about conscientiousness, how it's defined and measured, and what some of the empirical findings are.
There are also some really interesting theoretical papers. And it's interesting-- in particular, there's a theoretical paper by Conti and Heckman, which is a great paper for those of us in the social sciences to read. Because it's absolutely elegant in terms of the theory and his suggestions for the testing of theory. We do a little bit of what Conti and Heckman talked about in that article in this paper-- or in this talk and in the paper that I've developed that goes with this talk. And his ideas fit very well with a model that I'll be talking about that we developed a few years back.
So the impetus for this kind of work really comes from the fact that there's been a lot of interest in the economy in recent years given the economic downturn, the recession of 2008, the slow recovery that we've experienced since the recession occurred, problems of income inequality, and just a big focus and interest for policymakers about economic circumstances of families.
There's also a lot of interest in this-- what I call the stickiness of social and economic circumstances across generations. If you're born poor, there's a very high probability that you will grow up to be poor and that your children will be poor. If you're born well-to-do, there's a very high probability that your children will grow up to be well-to-do. And that will continue across generations.
There's also increasing interest in the pathways through which SES and human development are interrelated. We know that that socioeconomic status is related to lots of characteristics of human beings, both physical health and emotional health, and so on. But we don't always have a good understanding of why that is the case.
But in this particular talk, I'm going to be especially focused on this notion of stickiness. Why is it that there's cumulative advantages across time in some families and cumulative disadvantage in other families? And in thinking about the issue and in reading that special section in Developmental Psychology, it occurred to me that conscientiousness may play a key role in why it is that, particularly in more well-to-do families, that advantage continues from one generation to another.
The importance of conscientiousness-- conscientious people tend to work hard. They demonstrate self-control, set goals. They persevere in response to challenges. They tend to be responsible.
And it turns out that people with those kinds of characteristics appear to do better economically, socially, and emotionally than people without those characteristics. They also tend to live longer. They are healthier, so much healthier that they experience lower rates of mortality.
So the question that I ask in this particular paper, in this particular talk, is whether or not conscientiousness helps to account for the cumulative advantage or disadvantage that people experience as a result of the economic circumstances into which they are born, economic and social circumstances.
By the way, how long is this talk supposed to go on?
PRESENTER: It's scheduled for an hour. [INAUDIBLE]
RAND CONGER: OK, great, for an hour. OK, good. Because I could go from 15 minutes to three hours, so I'll do the one-hour version.
Well, if we're going to understand something about conscientiousness and how it has these good effects, it's important to understand where it comes from. How is it that some people are more conscientious than other people? And it turns out that self-regulation and self-control are central aspects of conscientiousness-- the ability to focus attention, to change attention as needed by the demands of the situation, and so on, the ability to control one's emotions appropriately in response to stress, and so on. And so self-regulation and self-control are central aspects of conscientiousness.
And it turns out that there's evidence that positive parenting characteristics-- warmth and support from parents-- seems to be associated in younger children with greater self-control. So Eisenberg and her colleagues have suggested that parenting may play an important role in the development of conscientiousness through its impact on self-regulation earlier in life.
That's the wrong direction. Excuse me. I thought I was hitting the laser pointer. Still not there.
So if it's the case that-- and there is evidence that positive parenting predicts greater self-control-- that raises the question of the context of parenting. If better parenting increases the probability that a person will grow up to be conscientious or a more self-controlled individual, where did that parenting come from?
And that takes us all the way back to SES. Could it be that SES affects the types of parenting that influence self-control and eventually personality development? So that's the starting place for this paper.
Well, in economics, there's something called the investment model, or what we call the family investment model. And it's one way to think about how SES might influence what parents do in terms of their children. Now, in economics, of course, they're thinking primarily about differences in income. So family income is associated with a range of investments, including living in better neighborhoods. If I have more money, I can live in a better neighborhood. Better housing, more adequate medical care, better nutrition, educational opportunities, and so on. If my child needs tutoring, I can afford to buy them to tutoring. I can get a tutor for them, and so on.
And we know from earlier research that these kinds of investments lead to increased human capital, child confidence, greater social and economic success. We don't know exactly from the economics literature how that Occurs that's the black box from economics. And in particular, the investment model does not specifically address the role of personality and its consequences.
But a few years ago, a colleague of mine and I, Bret [? Donelan ?] and I, came up with a model that we called the interactionist perspective drawing on Magnuson's idea of interactionist theory. And what we've done is we've examined family SES, parental investments, and personality development as part of a dynamic process where one influences the other across time.
And it's consistent with Magnuson's idea of an interactionist theory with Sameroff's transactional approach to child development, and so on. And when conscientiousness is the personality attribute of interest, as you'll see in this paper, I think it suggests one of the mechanisms whereby you get this process of cumulative advantage across time and across generations.
So here's the first model that I'm going to look at today in this topic. And we know from a lot of studies that education-- and I'm going to be drawing on the Iowa study, the Family Transitions Project. I started in 1991 when the adolescents in the Family Transitions project were about 14, 15 years old.
And we know from earlier research that education influences-- has a fairly robust influence in a number of studies on parenting investments in terms of things like support, consistent discipline, warmth, involvement, and so on. And we know that income from the investment literature, from the economic model on investments, we know that income is associated with the kinds of material investments that I talked about.
We would call these pathways a part of a social causation process. So we know that the social causation pathways do exist. We see these kinds of relationships between education, income, and investments.
Now, in this particular study, some people have taken education and income, and they'll use them as indicators of a single SES, Socioeconomic Status construct. But in reality, what happens in contemporary society, anyway, is that education typically is the forerunner of one's eventual income. People with greater education do better in terms of their income.
It's part of what your university president says when she goes to the legislature and asks for more money. People are going to do better economically, and we're going to have a more robust economic system if people have more education and better education. And the nice part about it when she says it it's actually true.
So as one sociologist said, education is the canonical variant in terms of SES. So in this particular study, these education reported by the parents, the original G1, the first generation parents in this study, that occurred typically about-- it was completed about 15 to 20 years before 1991. In other words, their total years of education had occurred quite a few years before. And we said that it would predict their income in the past year.
So we separate education and income for two different reasons. One, we think that based on previous research, that education will have a bigger impact on investments involving more effective parenting practices. Income will have a bigger impact on material investments, the things you can afford to give your child in a material sense, like living in a better neighborhood, and so on.
But the important point is that we predicted-- in this interactionist model, we predicted that both kinds of parenting investments would predict adolescent conscientiousness. So we measured conscientiousness in 1994 when these adolescents were about 18 years old, the last year of high school.
So all of this is part of the typical social causation process. And notice we're not looking at family stress now. We're really looking at family advantage. Higher SES in the form of education and income will influence parenting investments, expected to increase the development of self-control and similar kinds of characteristics of conscientiousness.
So for Eisenberg and her colleagues, they argue that self-control is the primary aspect of conscientiousness that's really important. We have earlier studies showing that parenting investments increase self-control. So we simply extrapolate and say that these kinds of investments are going to increase self-control and conscientiousness in general.
So that's the social causation aspect of the interactionist model. Now, what makes it more interesting is that these characteristics, these positive characteristics in terms of conscientiousness, are then expected to increase the likelihood that this adolescent is going to do well when they become an adult. This would be at age 25. This is at age 18.
They're going to do better in terms of completed years of education. They're going to do better in terms of the income that they earn. And that will feed back. Now, that would be called a social selection process. So now we've got social causation, social selection, and then social causation again.
Because we expect that these good things that are more likely to occur based on these personality advantages associated with earlier SES will create more good things in terms of environment, the SES environment. And that will feed back to increased conscientiousness across time.
So the actual success that the individual has based on these kinds of conscientious characteristics should increase their level of conscientiousness over time. It's kind of like if you think about a kid in school. If you give him successes, then they, over time, come to believe that they can be more successful. And so you're reinforcing their beliefs and their values about success in school.
So that's the first part of the model that I'm going to look at today. Then the next part of the model that I'm going to look at in the second hour-- no. We're going to look at the next generation of kids. We're going to look at G3.
Now, notice that this part of this model is exactly what we had over here. So in adolescence, education and income of the parents influence their parenting behaviors and influence the development of conscientiousness by the adolescents. Now, when we asked the same thing of these kids that we had followed for some 20 years-- they've grown up and had children of their own. And that's the third generation, G3.
And we ask exactly the same question. Will we replicate what we find with the G1 and G2 generation during adolescence? Will we replicate it when we go out to this third-generation child? And these kids averaged about 10 years of age in this study. And I'll explain that more later. But G2 education influencing their income, influencing the kinds of parenting investments they make in their child, and will that influence G3's development of conscientiousness?
Now, we have some important advantages here. Now we can control for parent conscientiousness before they became a parent, so before they could be influenced by the child's characteristics. So this is when they're 18 years old. And out here, they're in their early 30s, the G2s. And the G3s were about 10 years old.
Now we can control. If this is entirely a genetic phenomenon, none of these social factors, these social causation factors, should play any role whatsoever. It should just be you've got a conscientious parent, you're going to have a conscientious child. So we can test to see to what degree this might be a genetic connection.
We can also test to see whether or not the influence of education and income on these parenting investments-- is it because of education or income, or is it simply because of the parenting investments that G2 experience as a child? We know that kids tend to parent like their parents to some degree.
So maybe this link from conscientiousness to education down to parenting investments-- maybe none of that actually makes a bit of difference. Maybe it's just if you had good parents, you're going to be a good parent. And that's the whole story, and all this other mumbo jumbo doesn't really make any difference.
So I'm going to test those two models in this particular set of analyses. There is supposed to be a slide here that says I'm going to test those two models. But I think that slide comes later where it was supposed to come right here. So anyway, you know I'm going to test those two models. And I'm going to test them using data from the Family Transitions Project.
Now let me give you a little bit more information about the project itself. In 1989, we started the Family Transitions Project in rural Iowa. And it developed out of the economic downturn in agriculture in the Upper Midwest.
The downturn was absolutely dramatic during the 1980s. Every week-- I was at Iowa State University at that time. And every week in the newspapers, you'd read about farmers being thrown off the land. Every county-- and there were 99 counties in Iowa, they were small counties-- every county and every county courthouse, every week, they would put up white crosses in the courthouse lawns in the various counties to show how many farms had been thrown out of business, had been taken back by the bank in those counties in any given week. And those crosses were really building up.
You'd read about farmers shooting themselves because they went bankrupt. You'd read about bankers shooting themselves because they could no longer stand to foreclose on all these farmers. But people from the outside often don't realize the farmers and the bankers-- the farmer and the banker probably grew up together. They probably went to the same schools. They maybe married one another's sisters. And they knew each other well.
And now the fabric of their existence was being entirely torn apart. So this guy that I've known for 35 years and has been my friend, and we do all kinds of things together-- I've got to go tell him that I'm not going to give him the money he needs to continue to operate his farm this next week or this next year. I've got to do that to my old friend. And for some of the bankers, they couldn't stand it. And for some of the farmers, they couldn't stand it, either. And it was really terrible and tragic time.
We convinced NIMH back then that we should really study this and study how these families were coping with this degree of disturbance in their lives. So they gave us the money. It was very interesting, actually. A group from NIMH came out, and we showed them some videos of some of these families talking about their experiences. And of course, we had to mask the screens, because we couldn't reveal the identity of any families. These were just pretest families.
And these actually two or three, four administrators from NIMH, they wanted to know if this was really important. Because all they could think about when they thought about Iowa was these pristine farms with roses, and nice fences, and well-painted silos, and so on. We played some of these tapes of these people talking about their experiences. We got through the first couple of tapes, and these guys said, that's it. We don't need to see anymore. We understand now how bad the situation is. And they gave us the money.
So in 1989, we launched this study. And the adolescent portion of the study lasted in 1989 from 7th grade to 12th grade in 1994. There were a total of 556 adolescents. There were originally 400 and some in two-parent families, and then we added a supplementary sample of single-parent families to diversify the sample that we started. So we had a total of 556 of these generation 2 adolescents, all in rural Iowa.
And then in 1995, once they completed high school in '94, we continued to follow the original G2 cohort of adolescents as they moved into adulthood. And we followed most of them through several romantic partners. But almost all of them were married by their mid to late 20s. And many of them had had a child, the G3 child in the study. So we also studied the children of the children.
So the G2s were followed from an average of 12 years of age to 34 years old by 2010. And that's where I only do analyses with data out to 2010. We've collected some data since then. But over a 20-year period, we had 91% retention.
There's a couple of keys to the retention. One is you need to make them feel like it's their study, not your study. Another thing-- and probably more important-- is you need to pay them. And that helps.
But these retention figures and these kinds of longitudinal studies are a little bit misleading. Because people typically don't drop out completely. If you're going out-- let's say during adolescence, we went out every year. During adulthood, we've gone out every other year. They really don't drop-- most of them never drop out completely.
They might miss a year, or they might miss a couple of years in the series. But if we just looked at how many people were originally in the study at some time came back into this study, it would be more like 95%. So the retention figures are actually a little bit higher. There's a total of 350 G3 children back in 2010 from age 2 to age 17. So quite a number of G3 children.
Now, in the present analyses, I look at data from 1991, mean age when the kids were 14 years. I think that's really more like-- I think that's really more like 15, it's 14 to 15, out to 2010, between 33 and 34. Now, for the G1 parents and G2 cohort members back when they were adolescents, 19% of the G1 parents were single parents, so a single mother household. And almost always, it was a single mother household.
We focus on the G2 adults with a romantic partner. And that was $347. Now, between 80% and 90% of them had a romantic partner at any given wave of data collection. So by far, the majority had a romantic partner, and most of them were married.
And we focus on those folks because we want to get an informant report of personality. Most personality studies, you give somebody an obscenely long instrument of 300,000, 700,000 items that they fill out, and from that, you assess their personality.
But there is some data to suggest that if you get an assessment of personality by someone who knows the person well rather than from the focal individual themselves, you actually get better information in terms of predicting to other life outcomes.
So if I-- and in our own data, we know that if I ask Mom and Dad-- when the G2 adolescent was in 12th grade, if I ask Mom and Dad about their personality characteristics, those Mom and Dad reports predict much better to later life circumstances in terms of education, income, even romantic relationship quality, than self-reports.
And there are other studies that show the same kind of phenomenon. If you're predicting the future life outcomes, an informant report by someone who knows the person well actually predicts better, has better predictive validity than a self-report.
I think it's kind of like videotapes. You never see yourself on a videotape the same way somebody else would see you. And sometimes, we just don't see ourselves as clearly as other people see it.
Then G3 is following the same idea. G3 personality was assessed-- in terms of conscientiousness was assessed in 2007 and in 2010 by reports from the G2s and from their romantic partners. And there, we have an n of 282 that had that information.
We didn't start assessing personality with the G3s until they were six years old. Because we thought that was the earliest age at which we could get anything that would look like reasonable data. And the 12% of the families in which the G3 kids live were single parent families.
Measures education for G1, years of completed schooling by 1991, G2 education. There we looked at-- it in terms of markers of educational completion. So 1 would be less than high school all the way out to 6, professional degree or a PhD. G1 and G2 income were measured by income-to-needs ratio. That's total family income divided by the poverty level for a family of that size.
In the economics literature, this is the preferred index. Actually, if you estimate per capita income, which is just the total income divided by the total number of people in the family, it correlates about 0.9 to 0.95 with income-to-needs ratio, and it's a whole lot easier to compute.
G1 and G2 parenting investments-- those were based on observed positive and negative parenting in the family. And for that, we did videotaped interactions between family members. And based on those videotapes in the home-- all of our data, by the way, were collected in the home.
We go out to the homes of the families. We did videos of them discussing things, and then we rated those videos back at the lab for aspects of positive and negative parenting-- low hostility, high warmth, and support, consistent managerial practices.
The material investments index-- and I can give you more information about this if we have time and there's interest. We used material investments involving does the home have-- does it seem like a good learning environment. Is it structurally sound? What's the neighborhood like, and so on-- things that say that the parents are making material investments in the welfare of their child.
Conscientiousness-- again, we used other report, informant report because of the demonstrated improvement in predictive power. We used the multi-dimensional personality questionnaire, intelligence questionnaire. In 1994, we used that questionnaire to get a parent report on the subscale, the achievement subscale on the control subscale.
Achievement subscale says I'm ambitious and I want to get ahead, basically. The control subscale says I'm able to plan, I'm able to focus my attention on the things that I want to accomplish. I set goals, and so on.
So we used the Tellegen version there and the parent report in '94. People hated it. They just hated completing it. Because it's a series of vignettes, and you have to rate against that vignette. And so our interviewers came back and they said they hate it.
And so what we did-- and if I had to do it again, I wouldn't. It was foolish. But as you'll see, it worked out OK. We took the vignette-based Tellegen measure, turned it into a closed-ended questionnaire more suitable for survey research, and developed what we call the Iowa Personality Questionnaire, which is just a variant of MPQ and has the same achievement and control subscales.
So in 2005, when G2 was about 27 years old, the romantic partner reported on G2's achievement and control. And then in 2007 and 2010, the G2 and romantic partner, if there was a romantic partner in the home, completed the same reports on G3 achievement and control using the same things we used with G2 in 2005.
Keep in mind that these G3s are coming into the study each year once they hit the age at which we take them into the study, a minimum of 18 months of age. And so in 2007, we could have G3s who are anywhere from two years old up to about 17 years old.
So the two three four or five-year-olds, we're not getting the personality measure from them. We don't get that until they're six. So that's why we have-- we don't have a full 350 G3s with the personality measure for these analyses.
Ah, there we go, study result. So the first set of analyses focused on the period of adolescence and predictions to the G2 adult conscientiousness. And then the second set of analyses focus on predictions to G3 conscientiousness. Are there any questions so far about measures or study?
So just to remind you, here's the model. Now, you'll notice that the model also includes-- I said we're basically predicting that education will lead to parenting investments will lead to conscientiousness, and that will lead to greater education. But notice that we also have this direct path. And the direct path is there because we know that there is a strong relationship between parent and child education.
And there could be a lot of other factors that are influencing that, like relatives who also are highly educated and are influencing the child. We know also that there's a robust association between income across generations. So we leave that path in.
These are the structural parameters-- the where you are in the society, in lots of different ways, influences what you become in addition to what's going on in the family. So these are the extra-familial structural parameters that may be influencing what's happening to these kids.
So here's the standardized factor loadings for the various measures based on a confirmatory factor analysis for model 1, the model from adolescent to adult conscientiousness. And all of the factors loadings-- the factor loadings of 1 are for those variables where it's just a manifest variable, one variable estimating the construct.
The others are the factor loadings for the individual indicators for latent constructs, like parenting and investments. All of the factor loadings were positive. All were statistically significant and in the expected direction.
So I've heard statisticians say, well, that's unacceptable. 0.55 to 0.59 would be too low. Unless it's above 0.8, I don't care about it. I've heard others argue-- Nesselrode, for example, wrote a wonderful paper one time where he said you don't want them to be real high. Because if they're real high, you might as well just put them all together into a single measure.
So optimistically, you say, well, I'm tapping different dimensions of parity and investments. But I'll let the statisticians worry about that. They're significant in the right direction.
So here's the correlations for model 1. And you'll notice a couple of things right away. Notice that G2 adolescent conscientiousness-- all these correlations, by the way, are statistically significant. G2 adolescent conscientiousness-- 1994 at 18 years old, all of the predictors up to that point significantly predict-- at a 0 order level, significantly predict G2 adolescent conscientiousness.
G1 income predicts adolescent conscientiousness. G1 education predicts adolescent conscientiousness. And material and parenting investments predict adolescent conscientiousness.
So now the question is going to be is it really the case that these direct effects of SES, income and education, go away, and those effects are explained by the parenting investments in the model? Notice also, G2 adult conscientiousness is significantly related to all of the prior variables in the model, all the way back to G1 income, all the way back to G1 education still showing effects of material investments, still showing effects of parenting investments.
Now, if the mediational aspects of the model are correct, those direct effects should go away. And these impacts should be mediated through the other elements in the model. And here we go.
So a nice predictive path-- these are all standardized regression coefficients. The models fit reasonably well. Education predicts parenting investments, 0.50, as we predicted that it would. But it also predicts material investments, interestingly enough. And it predicts G1 income, as we expected.
Income predicts material investments, as we thought it would. But it also predicts parenting investment. Now, investment theory wouldn't necessarily predict that. Investment theory predicts-- it's going to influence how the family can invest materially in the child. There's nothing in investment theory by itself that says it's going to predict these other kind of-- warmth, support, consistency of management. But it does.
There's a whole-- I think there are lots of ways to explain these effects. But for our purposes today, the important thing is that SES does predict-- education predicts slightly better than income. Both parenting investments and material investments predict adolescent conscientiousness, as we expected.
And adolescent conscientiousness predicts well to G2 education. And G2 adolescent conscientiousness predicts well to income, some-- how long is that-- six, nine years later. This is personality at 18 predicting education and income nine years later. So pretty good. Pretty good.
However, notice education still predicts to education, G1 education to G2 education. But the magnitude of the association is about half what it was as a zero order correlation. So the mediation aspect of the model is supported in the sense that it really drops the direct connection between G1 education and G2 education.
Also interesting, emotional support-- good parenting practices also have a direct impact on educational success, even above and beyond these individual differences. So we're still getting these social effects in addition to the individual difference effect. Both of them are proving to be important.
Now, income drops out. There's no longer a direct path to income. So all of the direct path, the zero order association between G1 income and G2 income, is accounted for by the influence or the association of income with both types of parenting investments and with adolescent conscientiousness.
Interestingly, material investments also predict directly to G2 income. And I don't know how to explain that, to tell you the truth, other than maybe you get used to having a little bit better standard of living, and you aspire to maintain a little bit better standard of living once you get older. But I would not have predicted that, and I don't know exactly how to explain it.
G2 education-- now, most of them had completed their years of education by 21, 22, something like that. And this is at age 25. So their earlier-- for the most part, earlier educational accomplishments predict to their income. And both income and education predict relative increases in conscientiousness over time.
So we're getting this selection effect. Kids with certain kinds of characteristics do better. And doing better increases the magnitude of those individual differences.
Now, notice there's no direct effect from adolescent conscientiousness to adult conscientiousness.
AUDIENCE: Not a temperament.
RAND CONGER: Pardon me?
AUDIENCE: Not a stable temperament.
RAND CONGER: Well, there's not one single personality theorist in the world who would agree with that conclusion. And it may be that we have been so conservative in our measurement.
Notice now, this is adolescent conscientiousness assessed by the G1 parents predicting to adult conscientiousness of the same person estimated by a romantic partner. Now, the zero order association between these two things was 0.27. But it goes completely away.
I suspect if we had used-- if we do self-reports of G2 in both places, we probably would still have a significant relationship. The zero order association for self-reports is probably about 0.4, something like that.
But it's an interesting thought. Is it the case that using just self-reports by a single individual overestimates the degree of stability, simply because dispositional characteristics lead you to think about yourself in the world in similar kinds of ways over time? Or is it that this is too conservative?
AUDIENCE: What you're adding is a source of error, when you think of it. Because you are changing the person--
RAND CONGER: Absolutely.
AUDIENCE: That you're estimating. And not only-- and then changing the relationship, too, of one of romantic partnership to one of parental authority, forgetting that you are changing [INAUDIBLE].
RAND CONGER: We're changing a lot in that. We're even changing a little bit the form of the measure. But for self-report using exactly the same measure, the zero order association between these two things in other studies is about 0.04. The zero order association here is 0.27. So it really doesn't drop as much as you might think.
OK. Let's go on. Let's go on and look at G3. So again, the first thing we're going to want to be concerned with is how well does this portion of the model replicate with that portion of the model for G2 during adolescence. Here again, factor loadings-- they're all significant. They're all in the right direction.
Associations-- interestingly, once again, you'll notice that G3 conscientiousness is positively related to all of the antecedent variables in the model. Even to G1 material investments significantly predict G3 conscientiousness, and G1 parenting investments significantly predict G3 conscientiousness. So there's a lot of continuity across generations in these families. And here's the final model.
So we get the significant relationship from G2 education to G2 income, as we did before. G2 income predicts material investments. Education predicts material investments. And material investments predicts G3 conscientiousness.
Now, unlike the prior model, G2 income did not predict G2 parenting investments. Education did. But this is significant at the 0.05 level only with direction predicted. Now, we predict the direction, so I think it's fair.
But my colleagues say it's too loose. But it is positive. It's significant. 0.05 with direction predicted and parenting investments predict G3 conscientiousness.
Now, notice, though, that G1 parenting investments predict G2 parenting investments more strongly-- 0.37 standardized regression coefficient-- than G2 education in 2003. I think that's fascinating, because all of the studies that we do of parent-child relationships omit what happened in the history of the parents that we're looking at. We just don't have this kind of prospective longitudinal data.
And yet if we can take that into account, it's having a major impact, a very significant impact. In my mind, we're still getting a significant pathway here. It's very much attenuated compared to the model where we don't have that earlier information about the history of the child, the history of G2 as an adolescent.
We don't get a similar-- we don't get a similar direct path from G1 material investments to G2 material investments. And I think in a way, that makes sense. Because you can't provide a higher standard of living here, no matter what happened to you here, unless you have the income to do it. So current income should really have a bigger impact on material investments than the earlier material investments of the last generation of families.
So all this makes sense. That's significant. That's significant. Now, this path, I think, is really interesting. Now we are getting a significant relationship between adolescent conscientiousness and 1994, predicting out a minimum of 13 years to G3 conscientiousness in 2007 and/or 2010, 0.26.
So even though we've changed the measures, we've changed the informants, we're still getting a significant path, which leads me to wonder about, well, maybe that last path we saw in the last model wasn't so severely attenuated as we thought it might be by changes in measurement.
My genetics colleague, Mike Stallings, tells me that that is exactly what you would expect based on twin studies in terms of the genetic inheritance from one parent to a child. They said it's exactly consistent with a genetic effect.
It could also be that this person is conscientious as an adolescent, they're conscientious as an adult, and G3 is learning about conscientiousness from modeling by the parent as well as it possibly being a genetic effect. It could be either. It could be both. It's more likely it's both.
But in my mind, the important thing is we see a reasonably good replication of the same series of relationships-- education, income influencing the kinds of parenting investments that increase the likelihood of the development of conscientiousness by the next generation child. So we've got it over two generations in the same family. Yes. OK.
Any questions so far? OK.
So general support for the interactionist model of cumulative advantage now. If your parents are higher in education, higher in income, they're more likely to invest in a child in ways that promote a personality trait conscientiousness that is associated with perseverance, associated with self-control, associated with setting standards, having ambitions, and so on and so forth.
Those ambitions, in turn, promote future social economic success for the child. And that either feeds back to increased conscientiousness by the adolescents when they become an adult, or it goes on to influence their children through those socioeconomic advantages.
So the findings are consistent with processes of both social causation and social selection. The results are consistent with arguments of socialization and genetic effects. And the results indicate that adolescent conscientiousness contributes in important ways to adult SES in a fashion that affects adult personality development.
Now, one of the problems with this current design is it's not a genetically informed research design. It's not a twin study or something like that. So in future research, it would be good to have genetic information. It would be nice. And I hope NIH will fund one-- a twin study or an adoption study or something like that-- that will, over time, include multiple generations. I wouldn't hold my breath, because they just hate funding studies for that long a period of time. But it would be a good thing to do.
But there are possible genetic main and interaction effects in all processes. And there is the potential importance of epigenetic processes. SES-- these SES and parenting factors may have epigenetic influences that affect the way genes are expressed across time.
Now, we did get some money from NIH to collect DNA from these families. And we are working with those data right now. But I can tell you, the genetic effects so far in interaction with aspects of this model are not powerful at all.
But we've just begun that process. And we'll know more about potential gene by environment interactions or genetic effects-- we'll know more about that in the coming years. Obviously, a replication with other samples is needed, including with more diverse populations. I'm pretty optimistic that this work will generalize to more diverse populations, though. Because our earlier research on socioeconomic influences, primarily involving stress processes, has replicated from the rural white families in Iowa. It's replicated with African American families in both urban and rural settings. it's replicated with Mexican-American families, and it's also replicated in other countries. So I think these are human beings, and the effects are going to be similar in lots of other populations.
So SES family process is an adult personality or outcomes from prior phases of development. Part of the stickiness that we see in SES across time may well be a result of the related family processes and the related aspects of personality development.
Preventive interventions could, should target the early development of positive character attributes. So whether you're talking about improving the SES of parents if they're in economic bad straits, whether you're talking about improving parenting practices, or whether through schools or counseling or other kinds of things, you're providing the kinds of warmth, support, and other positive aspects of an adult environment for children. All those things could be potential areas for preventive interventions.
And if you can do things that would increase child and adolescent traits such as conscientiousness, you should have a positive impact on their ability to do better financially, to do better in terms of educational attainment. And those things, in turn, may help them to become a more effective parent themselves when they grow to adulthood and have those kinds of responsibilities.
So with that, I'll take questions.
AUDIENCE: Can I go back to the lack of correlation in conscientiousness within the same individual, child and adult, versus the presence of a correlation between child and offspring? Within the same individual, child and adult, you have it being graded by parent in the first instance and a romantic partner in the second.
Then in the second instance, where you're correlating a parental conscientiousness in G2 with child or adolescent conscientiousness in G3, you're correlating different individuals. So the fact that the latter is reliable and the former isn't seems counterintuitive, because in the first instance, you're holding individual constant, and in the second instance, you're varying it.
But I rather suspect that it's a difference in the way one applies criteria of conscientiousness. Because in the latter situation where you do have a correlation, the commonality is it's both parents doing the rating. You're rating conscientiousness from a parental point of view with all its implications for generality and overall success of a child.
I suspect that when you are in a romantic relationship, how you view conscientiousness is rather different. And perhaps-- this is not autobiographical, so stop that. Personally self-reference. Let's put it that way. [INAUDIBLE] fashion. And that may be the key factor. It's the one place in which the perspective on conscientiousness changes.
RAND CONGER: That's a good point. So zero order relationship between adolescent conscientiousness as reported by parents and adult conscientiousness as reported by a romantic partner is 0.27. The same zero order correlation for adolescent conscientiousness reported by parents and child conscientiousness reported by the next set of parents in 0.34. So it's a little bit stronger in the zero order case.
It's also-- another thing--
AUDIENCE: But not that much. So it could just be that the relations with the other variables are so much stronger that they're sucking out the variance in the places where there's no correlation between the child and the adolescent-- the child and the adult within the same individual. The other variables just aren't sucking out as much variances.
RAND CONGER: Right. If you look here, these paths are 0.66 and 0.22. And if you look here, 0.17, 0.17. It's a little bit smaller. Yeah.
AUDIENCE: So I'm curious about your take on this stickiness point that you raised earlier and the concept of cumulative advantage, which as you know is often invoked as an explanation for equality. So the idea being that as individuals or groups have advantages over others, the inequalities that are associated with those advantages also grow.
And the question is, do you, given your data, view conscientiousness as a kind of barometer for inequality that is capable of magnifying some of these small differences over time in such a way that makes it difficult for individuals in groups who are behind to, in effect, catch up?
RAND CONGER: I do. But keep in mind that conscientiousness is only part of the story. You're getting-- you're certainly getting significant-- and at least to education, a fairly robust path from conscientiousness to later education and later income. But you're also getting this direct path, so a social effect above and beyond the individual difference effect and a social effect here above and beyond an individual difference effect.
So in terms of cumulative advantage, the individual in the family with parents with a higher education is advantaged in several different ways, not just in terms of their personality characteristics. But certainly, they're advantaged here.
Susan Mayer in Chicago, she talks about the-- she doesn't think that income has much of an influence. But she thinks that what parents do who have higher incomes does make a difference in the sense that they're modeling for their children the kinds of persistent, effortful behaviors that the children also would need in order to succeed in the workplace, and that kids may be picking that up.
But if you take-- in my mind, I just imagine a difference. You have a child and-- well, I'll give you an example. Martha Cox and her colleagues have a study of rural families in-- I think it's in South Carolina and in Pennsylvania. And they're looking at-- do you remember the name of the study, Gary?
AUDIENCE: North Carolina and Pennsylvania.
RAND CONGER: It is North Carolina and Pennsylvania. And they were trying to find-- what they were trying to do is they were trying to get a full distribution of socioeconomic status across these families from both states. And they were trying to get the full distribution for white families and African-American families.
They couldn't do it. The African-American families didn't even come up to the mid-level income and education of the white families. They were so restricted in terms of opportunity, particularly in the South, if I remember it correctly.
A kid in those families-- they're blocked. In some ways, for a lot of them, it doesn't make much difference what the parents do, because they're not going to associate whatever it is the parents are doing with much success.
Now, that's an overstatement in a way. But when you can't even-- when, for a particular ethnic group, you just can't even find the distribution that would allow you to compare on socioeconomic status, you know that there is some major blockages that this child is exposed to that will be very hard for them to overcome.
The same thing happened with the neighborhood study in Chicago. Rob Samson, and I can't remember who all, was involved in that. They were trying to find African-American and white families living in similarly advantaged or disadvantaged neighborhoods. They couldn't do it, because they couldn't find African-American families living in neighborhoods as well-to-do as the white families. And they couldn't find white families living in neighborhoods as disadvantaged as the African-American families.
So there is-- it may well be the case that this development of conscientiousness, to a significant degree, is influenced by SES and by what parents do. But if parents are in an ecological niche whereby they can't promote their own success or the success of their children in the same kind of way, that also would restrict the development of conscientiousness and the ability of the child to build on what their parents have done.
So it's not just-- I would hate for a conservative commentator to say, well, if you just made all these kids conscientious, they'd do just fine, because it's not necessarily the case. But it wouldn't hurt. It's one aspect of the puzzle.
And you could give everybody more money. But if the kids didn't become conscientious, it could limit what that money could do. So I think that one of the nice things about a process model like this is it shows you several different points of potential intervention. And it would seem to me that interventions at all those levels that are associated with later success and well-being would be worth implementing.
AUDIENCE: Thank you so much for the talk. I found it really interesting. I wanted to ask about how we're thinking about education and income in the model. Because in addition to educational attainment and income at each of the generations being measures of socioeconomic status and context, I wondered if you would also be open to thinking about them as contexts of socialization in terms of the development of conscientiousness at those stages.
So for example, with the G2 education income, in addition to serving as a marker of socioeconomic status for that generation, could it also be-- I guess I'm thinking of the G2 education income pathways directly to the adult conscientiousness for G2 in 2005, so the 0.26 and 0.22 coefficients. Could those effects be interpreted as measures of socialization within the workplace or within the educational systems that would also affect conscientiousness?
RAND CONGER: And of course, that's the idea behind the model, that these conscientious people are more likely to engage in behaviors that lead to greater educational attainment and to greater economic success. And those experiences themselves reinforce the attributes that promote them by promoting later conscientiousness, and above and beyond their level of conscientiousness at an earlier point in time. So you are getting relative increases in conscientiousness across them.
AUDIENCE: Thank you for your talk. I really appreciate it. I grew up in Iowa and left the year that you study started. And so I actually appreciated the picture you drew, the importance of the real effect it had on families, on my relatives. I know this story quite well.
I had a question, maybe mixing a couple questions that were asked, in where you ended your talk talking about your expectation that the pattern of results should hold in different populations that you might look at, different ethnic groups that you might look at. And if it did, looking at the same types of variables, the latent concept, do you think-- when it comes to parental investment, for example, even if that level mattered, do you think the actual type of investments or the way that that manifests itself would be the same across racial and ethnic groups? So are parents actually doing the same things, or [INAUDIBLE]?
RAND CONGER: That's a great question. And of course, it's a topic of a lot of debate in the literature. But maybe what works if you're living in a very disadvantaged neighborhood-- it's dangerous, and so on-- then maybe the kinds of parenting practices that you engage in, the successful ones would be different than what would be the case in a suburban, upper middle class neighborhood.
I have colleagues who have used-- we have something called the Family and Community Health Study. And when I look at those data, I find that the same parenting practices that we saw with the rural, white families seem to have the same beneficial impacts with these African-American families living mostly in suburban, rural areas.
But some of my colleagues argue that that's not true, looking at the same data. And they argue that for the more disadvantaged families, you don't necessarily get the same kinds of effects. So more punitive parenting behaviors, for example, presumably for the more disadvantaged families don't have the negative effects that those behaviors would have for the more well-to-do families.
I have a couple of thoughts about that. And I have this good friend of mine, Wayne Ford in Des Moines, Iowa, who's a community organizer. And we used to argue about this all the time.
Some people will argue that the more punitive behavior by parents in these more restricted environs, more disadvantaged environments, aren't harmful in the sense that they don't appear to promote externalizing behaviors. When I analyze the data, they do. When other people analyze the data, they don't. And I've never really gone back to look carefully at why that's the case. Maybe there's some kind of magic mojo going on.
But they never promote anything, either. If you look at punitive behaviors, they may not be promoting an externalizing behavior, but they're not promoting success. They're not promoting competencies. They're not doing anything good. And they potentially could be doing something harmful.
Let's say, for example, that I'm a kid growing up in a significantly disadvantaged family. And my family, because it's kind of a dangerous environment, they feel like they have to be more-- they have to use more corporal punishment in order to keep me in line.
Well, now when I grow up, it could be I become quite successful, and I'm living in a middle class environment someplace. And now I'm using those same kinds of corporal punishment techniques with my children, because that's what I was exposed to as a kid. And now maybe in this new environment, they do promote anti-social behavior.
So I can't see-- I've never seen any data that says those kinds of behaviors do something good. And I have seen data that say they can do harm, at m under some circumstances. So why use them? Why foster them? I think they can generate a next generation of parents that will be more maladaptive with their children than they otherwise could be.
Now, there's a restrictedness that make sense. So if you live in a very dangerous neighborhood-- and I had a graduate student at Iowa State where this was the case, an African-American fellow. His mother wouldn't let him go off the front porch when he was a young kid.
She said, no, you're not leaving, because there's so much that's dangerous going on in this neighborhood. I'm not going to let you leave. She didn't beat him up. She just didn't let him go off the porch. So it wasn't getting knocked around that led to his success. It was just being restrictive. And you can be that without being--
AUDIENCE: I just want to comment, because that reminded me what Jerry Kagan always say. So what matters is not about what practice it is. What matters is the child's interpretation of the practice in terms of reinforcing development outcome.
There was actually a study done by [INAUDIBLE] Chen from Penn State. He looked at physical punishment, how that influenced child outcome in two communities. One community have an active perception about it's not good to physically punish your child. Another community, it's supposed to be OK to beat up your child, smack your child. So they developed outcomes very different. When it's perceived that it's OK, acceptable, children don't show negative outcomes from physical punishment compared with the other community.
I also wanted to ask a followup question. So this is really fascinating data. I'm very excited to see this. I'm wondering that conscientiousness predicts income and education. What kind of variables could [INAUDIBLE] that prediction?
So what exactly makes conscientiousness so important to determining, later on, the child's-- is that self-control of assistance, determination of resilience? What kind of [INAUDIBLE]? Because also, when you think across ethnic groups, we know Asians, Chinese and Japanese, score lower in conscientiousness compared with [INAUDIBLE].
RAND CONGER: Is that right?
AUDIENCE: I didn't know of that.
AUDIENCE: But as a rule, they don't necessarily perform worse in terms of education or income. But obviously, you cannot translate directly from the group level to the individual level. But I'm just thinking about something probably is working to make this conscientiousness as a personality trait.
RAND CONGER: With all these models, then, of course that leads to-- you lead to a whole other set of questions. So what is it that gets from here to here? And so you might say, well, a more conscientious individual is going to study harder. So you could look at studying. How much time do they spend studying? How much time do they spend on schoolwork or something like that?
You could look at values about education. Perhaps a more conscientious individual thinks that education is more important, and so on. But there's a whole series of other kinds of variables, obviously, that we need to understand between this and this to really know what's going on there.
It's like the connection between conscientiousness and better health. Why is it that more conscientious people are healthier? Well, they are more likely to go to the doctor. They're more likely to pay attention to signs and symptoms and so on. And they're even more likely to be married, and their spouse is more likely to get them to go to the doctor. So there would be a number of mediating variables that would account for that connection.
AUDIENCE: We should probably stop, because I know people have to be other places. But one final question is did you look at-- it goes back to some of these earlier questions about potential interactions in these models. And the question is whether conscientiousness operates the same way for low income versus higher income people.
Because if it's just, well, conscientious is a good thing if you're in middle or higher income and you want to come up, but it doesn't have the same effect if you're very disadvantaged families, then it suggests a different policy or intervention approach. Or the opposite, with conscientiousness promoting conscientiousness in low income families, you get more bang for you buck, so to speak, than you would if you focused in upper income families.
RAND CONGER: Well, some of these families, especially because they just been through some economic hard times, were fairly low income. But we did look at interactions between conscientiousness and education and conscientiousness and income. We didn't find any significant interaction. So it looks like conscientiousness has its effects across the different levels of education and income.
Now, interestingly, we also looked at aggressiveness, and it does interact with those variables. It means different things for people from different backgrounds.
AUDIENCE: Well, thank you so much.
[APPLAUSE]
Rand Conger, Distinguished Professor of Human Development and Family Studies at UC Davis, gives the annual Ricciuti Lecture Sept. 28, 2015. Sponsored by the Department of Human Development.