Research Roundup: A Surprising Benefit of Upskilling, Why Goals Can Backfire, and More

New research can transform the way you think about leadership, team dynamics, and business growth. This roundup, drawn from our May–June magazine issue, distills the most compelling findings from academics, consulting firms, and beyond. In this article you’ll learn: Why upskilling employees can help their managers focus on strategic priorities. How to get the biggest benefits out of your breaks. When using fewer points in your argument actually makes you more persuasive. Where relevant, we’ve included links to deeper coverage on HBR.org so you can dig into the findings that matter most to you. Upskilling Employees Frees Their Managers for More-Strategic Work Organizations routinely ask whether employee training “pays off” by boosting productivity. New research suggests that when firms track the results of such programs, they often look for efficiency in the wrong places. Researchers followed a Colombian government agency that had randomly assigned 12% of its frontline staffers to attend a 120-hour program on legal analysis, writing, time management, and basic IT skills. Using weekly goal-attainment scores before and after the program, the researchers found that four to six months later the performance of workers who’d participated had risen by about 10%. Their untrained peers, in contrast, showed no significant change in performance. However, the most meaningful impact of the training appeared to have happened one level up. Using email metadata to measure who communicated with whom, the researchers found that the trained workers had sharply reduced the number of messages they sent their managers. The untrained workers, in comparison, had reduced them far less. The training participants’ managers subsequently were more likely to achieve their own, more-strategic goals. Those who’d been most connected to trainees before the program enjoyed the largest productivity gains. In fact, the researchers say that the training’s largest benefit was to managers. Freeing up their time had ripple effects on the organization’s productivity, particularly as, at this agency, managers were heavily involved in valuable strategic tasks affecting their entire teams. The researchers believe that their study provides evidence that upskilling programs should be judged not only by how much faster or better trainees do their own jobs but also by how much less assistance they then require from the people above them. Especially in knowledge-intensive organizations where managers juggle oversight and strategic work, freeing up even a small share of their bandwidth can generate a lot of value. Go deeper: “Why Training Employees Pays Off Twice,” by Ben Rand. Even Small Slights Hurt Productivity Abusive managers clearly do harm to employees. But what impact do seemingly trivial managerial missteps have on workers? To answer that question, researchers partnered with a national retail chain in Israel to study how employees reacted when their managers let them down in a small way: by failing to deliver an expected birthday gift on time. At this company every employee with at least a year of tenure receives a gift and a card, hand-delivered by that person’s manager. The date of delivery is tracked digitally. Using data on 8,722 employees across 252 stores from 2018 to 2019, the researchers designated the people who’d received their gift within five days of their birthday (42% of workers) the control group. Those who’d received their gift in the month after their birthday formed the treatment group. The researchers noted that late gifts were not a reflection of strained manager-employee relationships or related to performance evaluations, demographics, tenure, or managerial style. They appeared to be accidental. Still, employees who received a gift late arrived at work late, left early, and took extra-long breaks more frequently the next month, something the researchers describe as “a cumulative form of revenge.” Because of this, they put in two fewer hours that month, on average. They also took 50% more sick days than their peers who received their gifts on time did. That increase translated into 216 lost workdays and $129,600 in annual losses for the chain. The researchers say that a late gift signaled a lack of care and respect, which made employees feel undervalued, causing them to lash out. Middle managers reacted the most strongly, probably because they identified more closely with the firm and had higher expectations for recognition and reciprocity. If something as minor as a belated birthday gift can measurably stifle performance, the researchers hold, everyday slights have a larger impact on the bottom line than most firms realize. Investing in training that makes managers more aware of their actions and more respectful to their reports may help prevent retributive losses. Overly Specific Goals Can Backfire Setting an explicit goal, like running a 5K or losing 10 pounds, can help you chart your progress and stay on track. But it can also cause you to miss out on more-effective ways to achieve results. In 13 studies with English speakers around the globe, researchers examined whether explicit goals caused people to stick with their original approach instead of switching to a better one. In one experiment 401 participants were asked to imagine they were trying a new meal plan to eat more healthfully and control calories. Half set a numeric caloric intake goal, while the other half did not. The groups were then told they’d chosen Meal Plan A and were given two low-calorie breakfast options. For lunch and dinner, they were offered the chance to switch to Meal Plan B, which had slightly more-healthful, lower-calorie meals and was better rated. Those who’d set a specific calorie goal were less likely to switch to Plan B, even though it was the superior plan. The same pattern emerged in other experiments involving investment scenarios: People who’d set an explicit earnings target were more likely to stick with their initial investment-fund choice, even when other funds that could serve their goals better were available. Across the experiments, setting a numeric goal—and making some progress toward it—tended to make participants perceive the initial method to be more effective and view the alternative as less so. But when they were prompted to reflect on the benefits of both the initial and alternative methods, that bias disappeared. The researchers write that this has implications for managers helping their employees reach goals. Rather than tracking progress from an initial plan, they should encourage employees to pair objectives with structured prompts to periodically reassess their approach. That may help them choose better, more-effective means of getting results. How to Benefit More from Breaks Lots of research shows that taking a breather can help you stay focused and prevent long-term burnout. But how you organize your tasks may be just as important as the break itself. Across five studies researchers examined what happened when people took a break between two tasks or in the middle of one long task. In one, 398 U.S. adults performed the same word-search activity. Half were told they were doing a single 10-word puzzle. The others were told they would do two separate five-word puzzles. After finding five words, all participants watched a short, relaxing video. They then reported how much they’d thought about the puzzles during the break and how stressed or anxious they felt. Those in the group tasked with one long puzzle thought about the task significantly more and expressed higher levels of stress and anxiety than the group assigned two shorter puzzles. Next the researchers tested performance. They ran the word-search experiment again, but this time they measured how quickly participants were able to find words after the break. Those in the group that solved two short puzzles performed significantly better after the break than those solving one long puzzle. This was largely due to less time ruminating about the puzzle during the break, the researchers say. Finishing the first puzzle provided a kind of closure that allowed participants to more fully unwind before their next assignment. The researchers also found that the type of break mattered: Only passive breaks, such as watching a scenic video of the Swiss Alps, drove better performance. Active breaks, like engaging with an interactive tour of Buckingham Palace, produced no performance gains after the break. Beyond just taking breaks—or encouraging your employees to do so—it’s important to categorize the work surrounding downtime, the researchers say. Treating tasks before and after breaks as distinct projects can make stepping away more restorative—and your return to work more productive. Simplify Your Arguments to Win People Over When you’re trying to gain support for an idea, it’s tempting to pile on justifications and address every possible objection. But new research shows that that approach isn’t as successful as you’d think. Across five experiments, researchers varied how many reasons a hypothetical person gave when arguing for a position. In one experiment, for example, 475 U.S. participants read an email from a fictitious community member about an upcoming ballot measure. The email contained either one or nine arguments in its favor. Participants were then asked to rate their attitude toward the measure, how motivated they believed the community member was to try to persuade them, and how expert they thought the member was. The people who were given a lot of arguments were more likely to consider the community member an expert, but they also found that person’s intent to persuade them much stronger, which made them suspicious. Statistically, the two effects largely canceled each other out. Nine arguments were no more persuasive than one was. That held true in additional experiments varying the number of arguments on topics such as trying a new coffee brand or rejecting measures to bring about universal healthcare. However, when the context made one perception more important than the other, the number of arguments mattered a lot. In later experiments participants who were told to “listen to experts” were won over by lots of rationales; those told to “avoid people with an agenda” were more persuaded by just a few. The researchers say that to build a convincing argument, you need to understand what your audience values. In expertise-driven settings, like technical decisions or risk reviews, a richer case can help. In contexts where audiences are wary of spin, like contentious internal changes or consumer marketing, a concise, well-chosen set of arguments may actually be more effective. LLMs Respond Differently in Different Languages A new study suggests that the language you use with generative AI can affect its answers. In it researchers analyzed more than 10,000 responses from two leading models: OpenAI’s GPT and Baidu’s Ernie. They asked the models questions about two foundational constructs in cultural psychology: social orientation (how people define themselves in relation to others) and cognitive style (how people habitually process information). The researchers used well-established cultural patterns as a reference point for their predictions. In the United States people tend to have an independent social orientation characterized by autonomy and uniqueness, along with an analytic cognitive style relying on logic and rules. In China people tend to have an interdependent social orientation emphasizing harmony and connectedness, along with a holistic cognitive style focused on context and open to contradiction. The researchers found that the generative AI models—trained on textual data that is inherently cultural—consistently exhibited those cultural tendencies when prompted in each language. When posed questions in Chinese both GPT and Ernie exhibited a more interdependent social orientation and holistic cognitive style. For example, both models attributed a person’s actions at a basketball camp to the context rather than to individual personality traits. These tendencies can have real-world business implications. When the researchers asked the models to recommend the most effective marketing slogans for fictional startups, the recommendations shifted simply because the prompt language had changed. The researchers also found that these tendencies aren’t necessarily fixed. When the generative AI was prompted in English to adopt the perspective of a Chinese person, its responses more closely resembled those it produced when prompted in Chinese. “Leaders should be aware that prompt language can quietly influence [generative] AI’s recommendations and, in turn, business decisions,” the researchers write. “Managers and executives, especially those who work with multilingual teams or global customers, should therefore treat language choice as a strategic decision.” Go deeper: “Research: LLMs Respond Differently in English and Chinese,” by Jackson G. Lu and Lu Doris Zhang. Making Games Easier Can Boost Revenue Designers of free-to-play games often ramp up their difficulty to get players to purchase add-ons that help them advance. But the opposite tactic may be more profitable. In an experiment with more than 300,000 global players over a 12-week period, researchers found that lowering a game’s difficulty significantly increased player engagement and retention, leading to an increase in long-term revenue. “Personalized Game Design for Improved User Retention and Monetization in Freemium Games,” by Eva Ascarza, Oded Netzer, and Julian Runge. The Best Soundtracks for Ads Using a song in an ad campaign can make a product more appealing—but not all songs work equally well. A new study shows that ads paired with songs featuring highly repetitive melodies generate the strongest consumer responses, both to the music and to the product, than the same ads set to more complex or less repetitive music. “The Impact of Melodic Repetition,”by Giovanni Luca Cascio Rizzo and Jonah A. Berger.

Research Roundup: A Surprising Benefit of Upskilling, Why Goals Can Backfire, and More

Summary.

Partner Center