Ask Again, and You Shall Possibly Receive

It sometimes seems that our evolutionary ancestors would be turning in their primitive Neanderthal graves if they knew how often we fail to take advantage of the awesome social abilities we’ve developed. Case in point: people don’t ask for help as much as they should. Despite our altruistic social norms and the ability to communicate via Snapchat, people are surprisingly content to lug 12 bags of groceries or study for a physics midterm all by themselves.

Why aren’t we more willing to seek help? One possibility is that people underestimate the chance help will be given. A new study led by Stanford’s Daniel Newark examines if and how this might happen in one specific type of situation — when a help-seeker makes a second request for help after an initial request is turned down.

Newark and his team believed there were two reasons people denied help would underestimate the chance they would receive it if they asked again. First, they speculated that help-seekers would underestimate how uncomfortable it is to turn down a second request. Studies have shown that people feel pressure to say “yes,” and such pressure is likely to be greater when they’ve already said “no.” But when somebody turns down your request for help, your rarely consider how the decision made them feel. So while a help-seeker sees an initial denial as a sign there’s little anguish over saying “no,” a person receiving a second request faces even more anguish due to the fact that they’ve already said “no” once.

The second reason we’re likely to be bearish about receiving help is that people tend to attribute the actions of others to unchangeable dispositional factors rather than the specific situation. So when somebody refuses to help, it’s assumed that they’re just not very generous, not that they had a good reason to say “no.” The result is that people will consistently underestimate the generosity of those who deny a request for help.

After an initial survey experiment found support for their hypotheses, the researchers sent participants to a college campus to ask people to fill out a 1-page survey (first request) and then mail a letter (second request). Prior to going out into the field participants predicted how likely potential helpers would be to consent to their requests. It turned out participants were quite good at predicting whether people would respond to the initial request. Participants thought 34% would consent, and in reality 33% consented. But as the researchers predicted, participants failed to foresee how people would respond to a second request after they denied the first request. Participants predicted an initial denial would lead to a “yes” only 18% of the time, but 43% of potential helpers ended up agreeing to the second request after turning down the first one.

Two follow-up experiments in which participants imagined being a helper or a help-seeker confirmed the initial findings. Once again, help-seekers underestimated the likelihood that a helper would agree to help after turning down an earlier request. In addition, the experiments provided support for the role of discomfort and dispositional attributions. After being told their initial request was turned down, help-seekers and helpers rated the helpers dispositional “helpfulness” and the discomfort helpers would feel from saying “no” to a second request. As predicted, help-seekers rated the helpers as less helpful and thought they would feel less discomfort.

One question that remains is whether to tendency to underestimate the likelihood of help differs across cultures. A Western culture that emphasizes individuality and self-sufficiency may foster the belief that others are unlikely to help. However, in a collectivist culture where being denied help is less common, an initial denial could lead somebody to believe that a second request would surely be granted.

It’s also worth considering whether underestimating the likelihood of help could be a good thing. In an individualistic society, such beliefs might encourage you to figure out how to take care of things on your own. Being denied help can also be a painful experience, and if such pain is significantly greater than the utility of being helped, in the long run being discouraged from asking for help could lead to higher well-being.

But for most people grinding their way through the day, the lesson from the study is obvious. If somebody refuses to do a favor for you, don’t hesitate to ask them again. Many downtrodden cavemen were booted from their tribes to create the social norms that make a potential helper more likely to say “yes.”

(cross-posted from The Inertia Trap)
Newark, D.A., Flynn, F.J., & Bohns, V.K. (2013). Once Bitten, Twice Shy
The Effect of a Past Refusal on Expectations of Future Compliance Social Psychological and Personality Science DOI: 10.1177/1948550613490967

Tony Bennett’s Not Corrupt, He Just Did His Job Very Very Poorly

Michael Petrilli and Andy Smarick took to Twitter today to try and explain why Tony Bennett did what he did (Petrilli’s longer thoughts are here.) The basic defense is that when Christel House Academy scored a ‘C’ it tipped off Bennett’s team that something was wrong with the system. An error was found regarding schools of a certain grade span, and the error was corrected, causing the grades of a handful of schools to change.

If that’s the case, then fine. But that doesn’t square with the story:

When Bennett requested a status update Sept. 14, his staff alerted him that the new school grade, a 3.50, was painfully close to an “A.” Then-deputy chief of staff Marcie Brown wrote that the state might not be able to “legally” change the cutoff for an “A.”

“We can revise the rule,” Bennett responded.

Over the next week, his top staff worked arduously to get Christel House its “A.” By Sept. 21, Christel House had jumped to a 3.75. Gubera resigned shortly afterward.

After the error was found, the formula seems to have been continuously revised until Christel received an ‘A.’ I’m willing to give Bennett the benefit of the doubt — I don’t think he was corrupt or attempting to give illegal favors. I think he was using Christel in a genuine good faith effort to create the best system. But by using Christel to calibrate the results Bennett didn’t create a system that rewards schools for reaching an objective standard of excellence, he created a system that rewards schools for being like Christel. That’s a bad way to create an accountability system.

Once the error was recognized there’s only one thing that should have happened. Bennett’s team should have met and decided how they thought it was best to grade the affected schools, without bothering to look at any individual school. Bennett has a team of trained professionals and statisticians and they should be objectively grading schools on what they believe to be the best mutually agreed upon metrics. They shouldn’t need past results to “check” their work. If they do, the system is not objective. If the federal government decided on a rating system for universities, but then tweaked it because Harvard and Yale only got a B, people might not suspect foul play, but they wouldn’t say it was a good and objective way to build the system.

Opponents of accountability have had to swallow a lot of bitter pills over the last few years (pills I’ve generally been happy to see them swallow.) But the saving grace has been that the accountability is objective. The system is blind to specific schools, and that means everybody is treated the same. Bennett sidestepped that standard in way that wasn’t appropriate for a public agency. He never should have been assuring good schools that they would receive good grades. The methodological merits of the system should have been able to stand on their own.

On Twitter Neerav Kingsland asked Smarick how we know that Christel is high performing, and Smarick replied that it had done well under previous grading systems. But the previous grading systems were flawed! That’s why a new one was being made. Even if Christel’s good performance in past systems was “real,” and even if you accept that Christel is objectively a great school, allowing a single school to influence an accountability system pushes you to a point where all objectivity starts to leak out. In this specific instance it may not have led to an inferior grading system, but using a single school for calibration is, to put in mildly, not a best practice. Thus Bennett’s sin wasn’t corruption, it was merely doing a really bad job.

Life Satisfaction Predicts Population Growth

From a new study by Michigan State’s Richard Lucas:

Subjective well-being (SWB) reflects an overall evaluation of the quality of a person’s life from his or her perspective. Although SWB is typically studied at the individual level, social scientists have become increasingly interested in the well-being of broader regions like cities, states, or nations. The current study examines the association between aggregate well-being and an important behavioral indicator of regional success: migration and population growth. Using life satisfaction data from over 2 million respondents, along with population data from 2000 to 2010, I show that U.S. counties with higher levels of life satisfaction grew at substantially faster rates than did counties with low life satisfaction. Supplemental analyses showed that this association was not due to regional differences in birth or death rates. Instead, counties with high life satisfaction experienced high levels of domestic migration. These results show the validity and utility of life satisfaction measures at the regional level.

Don’t Conflate Education Costs With Education Spending

In public policy there’s an important difference between between cutting costs and cutting spending. Cutting costs potentially allows you to have more of something. For example, in healthcare, if a cost-cutting technological advance allows you to get a procedure with 90% of the effectiveness for 50% of the cost, you could take the money you save, spend it on other health services, and potentially end up with more “health” than before the cost-cutting. An action that cut costs leads money to be spent more efficiently, and you can end up better off because of it.

Cutting spending is different because it reflects a decision to have less of something. If your overall healthcare budget is cut by 15%, you’re going to end up with less “health” than before. So while cutting the costs of a given service could allow you reap more overall benefits, cutting spending will leave you with fewer benefits.

Right now there’s a battle being fought to blur this distinction in both higher and K-12 education policy. For example, proponents of MOOCs view them as a way to cut costs and ultimately get more “learning.” If an MOOC can deliver 80% of the learning that occurs in a conventional class for 50% of the cost, those savings can be spent on other important things (scholarships, tutors, professors, etc.) The result is that students can eventually end up with 110% of the previous quantity of “learning.” On the other hand, opponents of MOOCs feel cornered by growing budget deficits and conservative state legislatures. As a result, they view all cost cutting as a means to cutting spending. In their eyes, even if MOOCs give you 80% of the learning for half the cost, you’re just going to be stuck with 80% of the learning. There’s no reinvestment of saved resources on its way, and that means any efficient cost cutting measure will lead to corresponding cuts in spending , and a corresponding decrease in learning.

There’s a similar debate that takes place in K-12 education. For example, proponents of charter schools that use blending learning might believe the model allows students to learn 90% of the necessary math in 70% of the time (or for 70% of the cost). The saved time could then be used to teach math in small groups, and the result is that students end up learning 110% of the necessary math. Others feel that such a scenario is a pipe dream. Even if the school manages to get 90% of the learning in 70% of the time, the remaining 30% will end up as non-profit or corporate “profits” and/or budget cuts. In the end students are left with 90% of the learning. As with higher education, the argument relies on the assumption that the savings from cutting costs are not spent on more learning. Or as Diane Ravitch lays it out (succinctly, as always):

The holy grail for corporate reformers is cost-cutting that produces profits. Their hope is that if schools replace teachers with technology, the districts save money, and the tech companies strike it rich.

Ravitch speaks for a movement that assumes certain cost savings are a trojan horse for cuts to spending on learning (and by “spending on learning” I mean spending that actually goes toward educational services, not the per-pupil allocations that show up as “spending” in a budget but could go toward “profits.”) The movement tends to oppose charter schools, education technology, and the weakening of teacher credentialing, initiatives that some believe could help cut costs in an efficient way. So the question is, will cutting costs lead to less spending on learning? Should we believe that no reinvestment will take place?

I think the answer is no. While the “cutting costs = cutting spending” argument is fairly compelling when it comes to higher education, there is little evidence to support it in K-12 education. One reason is that K-12 education has real accountability systems, whereas higher education has none. If MOOCs are being used to provide inferior learning opportunities, not only is it difficult to prove, there are no consequences if proof is found. We basically let colleges do whatever they want. Things are different in K-12 education. If a charter school provides an inferior education, it’s likely to get shut down. Perhaps the frequency of such closings is still suboptimal, but they are not the myth some would have you believe. In fact, a national charter school advocacy group has a large-scale initiative to close down more schools! That’s hardly the action of a sector bent on providing inferior learning in order to skim tax dollars. The bottom line is that if you cut spending on educational services, there will likely be consequences. Even truly malevolent actors won’t be able to cover their tracks by enrolling a student body whose special education population is 1.4 percentage points lower than the local district average.

K-12 education also has an extremely powerful constituency of parents and teachers that’s vigilant about fighting abuse. The same can’t be said for higher education, where states have routinely been able to cut spending. Furthermore, there’s not a lot of evidence that the supposed bad actors driving these cuts — chiefly corporate CEOs and conservative state-level politicians — have a desire to pale back spending on K-12 learning.  There’s a knee-jerk reaction to assume that the education policies of state-level politicians will mirror the ravenous cost-cutting appetite of the crazed Congressional GOP, but governors have to be responsible, and they generally understand that education is valuable. Take a look at this chart of state-level changes in education spending over the last five years:



Notice anything? Ok, that’s a trick question. You’re not supposed to notice anything because there’s not much to notice. No patterns jump out with regard to the political ideology of state legislatures or the market share of charter schools. Granted, every state had their own unique circumstances during the recession, but there’s nothing that would lead you to believe an identifiable group is using the idea of cost-cutting to slash spending on learning. Even budget-cutting addict Rick Scott wants more spending! Cynics might say he just wants to funnel more money into his cronies’ pockets, but even if that’s true his proposal is certainly not the type of thing you do if you’re scheming to spend less money on learning.

Yet despite a lack of empirical evidence, much of an entire movement’s argument remains built on the fact that any kind of efficient cost-cutting will lead to less spending on learning. The result is that much of what’s written becomes an exercise in making ad-hominem attacks and playing “Six Degrees of Some Crazy Tea Party Dude.” It may seem farfetched that Rocketship charter schools are only interested in teaching math more efficiently so they can raid education budgets, but if their funder’s funder’s friends are evil, maybe Rocketship really is up to no good.

That’s why every time certain entities are mentioned the name of 10 other loosely affiliated entities (all bad, of course) are mentioned along with them. (My favorite example is this utterly nondescript Diane Ravitch postIt’s literally one sentence of news about a Khan academy grant followed by 8 sentences of ad-hominem attacks against a dead woman.) It’s also why there’s a concerted effort to turn Bill Gates — a man who is unquestionably on Earth’s top 10 list of “People who have alleviated the most human suffering” — into the Koch brothers — seemingly selfish tycoons who have few qualms about harming others (or destroying the environment) in order to promote their own business interests. There are good reasons to disagree with Gates’ policies, but if your argument is built on his nefarious motivations, it may be time to re-examine your priors. (Or to put in another way, you may be derpy.)

But I digress. My point is that it’s bad to equate cutting costs with cutting spending because cutting costs could ultimately lead to more learning while cutting spending will not. But when unsubstantiated assumptions lead you to refuse to acknowledge that cutting costs can be a step toward smarter spending and more learning, you rule out those potential gains. To go back to the healthcare analogy, it’s like refusing to ever purchase a slightly inferior but much cheaper treatment from a certain group of doctors, with the result being that you can never efficiently reallocate spending and reach a higher level of “health.” That’s not to say people shouldn’t scrutinize big spending decisions on technology or charter schools, but when you repeatedly assume that a slew of potentially good things are bad, you’re eventually going to stop good things from happening.

To Change Behavior, Focus on a Single Situation

Atul Gawande has a great piece about the spread of innovation and the difficult process of changing deep-rooted behaviors. For example, while anesthesia became widely used in surgeries almost immediately after it was invented, it took decades before people consistently sterilized medical environments even though the danger of germs was well-known.

The surgeon J. M. T. Finney recalled that, when he was a trainee at Massachusetts General Hospital two decades later, hand washing was still perfunctory. Surgeons soaked their instruments in carbolic acid, but they continued to operate in black frock coats stiffened with the blood and viscera of previous operations—the badge of a busy practice. Instead of using fresh gauze as sponges, they reused sea sponges without sterilizing them.

Gawande goes on to describe the difficulty of training third-world birth attendants on best-practices that lower infant mortality. The good news is that meticulous but sustained in-person instruction appears to be making a difference.

Of course for most people, attempts to change behavior involve lower stakes. Eating healthier. Going to the gym. Calling your grandmother more often. But like the surgeon who must remember to constantly kill germs — when getting dressed, when taking out instruments, when entering the surgical room — somebody attempting to give up junk food can also feel as though constant vigilance is required. Watching TV, going to the water cooler, or meeting a friend at a bar can all lead to temptation. The result is that plans to change a certain behavior often involve a variety of situations in which people aim to be better.

This war-on-all-fronts strategy may seem promising, but a new study suggests it’s actually a bad idea. Researchers from the University of Utrecht wanted to find out if people would be more effective at changing their behavior if they made a single behavioral plan (technically called an “implementation intention”) rather than multiple plans. For example, if you’re trying to avoid checking Twitter, would it be more effective to make a single rule — e.g. “If I’m on a conference call with the Chicago office, I won’t open Twitter” — or, in addition to that rule, to also commit to not check Twitter upon waking up in the morning or getting into a taxi. The researchers reasoned that the multiple implementation intentions might “dilute” each other and lead to less behavioral change.

The study examined the unhealthy snacking behavior of participants who had previously signaled interest in reducing their snacking. Participants were divided into three groups. One group chose a single implementation intention that they wrote out in a format of “If [triggering cue], then [solution/altered behavior]” — for example, “If I watch Game of Thrones, I’ll eat only an apple.” A second group chose three implementation intentions, and a control group was simply instructed to think about healthy alternatives to unhealthy snacks. The participants then used a diary to track their snacking behavior over a three day period.

The results suggest that trying to change behavior through multiple plans is a fool’s errand. Relative to both the control and single-plan conditions, participants who made three plans reduced their snack consumption and calorie intake by a significantly smaller degree. In fact, relative to a baseline measure, participants in the three-plan condition consumed more calories. The single-plan participants performed better than those in the control condition, although both groups did quite well and the difference was not statistically significant.

A follow-up experiment supported the initial results and found evidence that there is a cognitive basis for the superiority of a single plan. Compared to participants who chose three situations in which they would resort to an alternative healthy behavior, participants who focused on a single situation were faster to recognize words that were part of the alternative behavior. Taken together, the experiments suggest that multiple plans do dilute one another and make a substantial change in behavior less likely.

The lesson is one that parents and teachers are familiar with. Take it slow. Go one step at a time. If you want to check Twitter less, focus on making breakfast instead of reaching for you phone when you first get out of bed, and focus only on that one change. Rome wasn’t built in a day, and the same can be said for improving your diet, whether it’s cutting the consumption of tasty but unhealthy snacks or entertaining but frivolous tweets.
Verhoeven, A.A.C., Adriaanse, M.A.A., de Ridder, D.T.D., de Vet, E., & Fennis, B.M. (2013). Less is more: The effect of multiple implementation intentions targeting unhealthy snacking habits European Journal of Social Psychology

Yes, Fox News Matters

From a new paper by Penn’s Susanna Dilliplane:

This study uses multiwave panel data from the 2008 presidential election to investigate the impact of partisan news exposure on changes in vote preferences over time. Overcoming key limitations of prior research, the analysis distinguishes among the potential effects originally delineated by Lazarsfeld and colleagues (1948): (1)activation—motivating partisans who initially say they are undecided or planning to defect to shift their vote back to their own party’s candidate; (2) conversion—motivating partisans to shift their vote to the opposing party’s candidate; and (3) reinforcement—strengthening partisans’ preference for their initial vote choice. The results reveal only modest evidence that partisan news reinforces existing vote preferences. Surprisingly, partisan news plays a more robust role motivating changes in vote choice: news slanted toward citizens’ own partisanship increased the odds of activation and decreased the odds of conversion, while news slanted away from citizens’ own partisanship proved a strong counterforce working in the opposite direction.

The Secret to Finland’s Teachers Is That It’s Easy to Decide to Become One

I’m not a big fan of drawing sweeping lessons from education systems in other countries — I’m partial to the Rick Hess school of “they’re just too damn different” — but there was something about Finnish teachers in this weekend’s NYT op-ed on testing that’s worth highlighting. If you’ve been living under a rock for the last 10 years, the standard trope is that other countries not only have better teacher training, they have higher standards that ensure only the best of the best end up in classrooms. Or as the NYT op-ed summarizes:

Perhaps most important, they set a high bar for entry into the teaching profession and make sure that the institutions that train teachers do it exceedingly well.

But then in the very next paragraph, there’s this:

In Finland, for example, teacher preparation programs are highly competitive and extremely challenging. (The programs are free to students and come with a living stipend.)

When it’s hard to become a teacher because the bar is so high, that’s a good thing. But when it’s hard to become a teacher because choosing to do so is a difficult decision, that’s a bad thing. In Finland preparation programs are not only free, they pay you money. That makes the decision to become a teacher relatively easy. Even if you’re so talented you have good offers to work at a bank or sell software, teaching is still an attractive option.

In the U.S. that’s not necessarily the case. People interested in becoming a teacher not only have the cost of graduate school tuition, they have the opportunity cost of not being able to earn money for a year. For talented young people who have other options these costs make is difficult to decide to be a teacher. A 22-year-old needs to be quite sure about something in order give up $50,000 (this assumes a functioning economy and $30k in salary + $20k in tuition ). The result is that unlike the Finnish teacher prep system, which is built to filter for people who will be great teachers, the American teacher prep system is built to filter for people who really want to be teachers. Any discussion of the differences in the U.S. and Finnish teaching force should take this into account. It’s not merely some vague notion of “prestige” or a lack of standardized testing that attracts better candidates to Finnish teaching jobs, the immediate financial prospects of training to be a teacher are also much more promising.

In fact, the lone U.S. initiative to replicate Finland’s success in recruiting top talent happens to be similar to the Finnish system from a financial standpoint. What initiative is that? Teach For America, of course. Just like Finnish students who are preparing to be teachers, first year TFA teachers end up with positive cash flow. There are a variety of things that attract people to TFA, but the you can’t ignore the fact that it eases the standard financial burdens associated with becoming a teacher.

None of this is to say that the U.S. should drop everything and focus on moving to a no-cost training system. At this point I think it’s best to stay the course and focus on improving training and professional development rather than launch harebrained schemes to get ivy leaguers to become teachers. But the reason I bring this up is to show that a higher bar often creates a lower ceiling. If you want to make becoming a teacher so financially and academically rigorous that you weed out all the weak candidates, you’re also going to make teaching less attractive to the strongest candidates. Finland is able to have a high bar and a high ceiling by using subsidies to eliminate the financial rigor and keep the academic rigor. Unfortunately, not only is that politically unfeasible in the U.S. (hooray for idiotic austerity!), it seems we’re closer to the opposite arrangement. Teacher prep programs have a big financial cost while their academic reputations are more akin to a rubber stamp. Has any non-teacher ever heard of somebody failing out of a teacher prep program?

The broader point is that improving the pool of teaching candidates is complicated, and many people who see it as a priority tend to have unserious proposals. For example, a legitimate budget-neutral solution would be to raise top salaries but lower average salaries by creating a more varied performance-based pay scale, then use the savings to make teacher prep programs free and with stipends. The improved financial arrangement as well as a chance to be a paid like a teaching superstar would attract a larger proportion of elite graduates. Of course this proposal is the opposite of what teachers want, but that’s as it should be. With the exception of across-the-board raises, incumbents and potential entrants generally value different things. So for all the talk of the need for it to be easier to choose to become a teacher, people generally aren’t willing to sacrifice anything to make it happen. It would seem that in this respect, nobody wants to be like Finland.