praise, gender in the workplace, and public speaking, in the latest from the Society's free Research Digest service (see www.researchdigest.org.uk/blog)

An 'incredibly' bad idea?

In Psychological Science

When you’ve done something good, or performed a task well, it feels great to get some praise for it. And parents and teachers, especially in Western cultures, are encouraged to dole out praise to children in an increasingly generous manner. A drawing might not just be ‘good’, it might be ‘incredible’. That song wasn’t just ‘beautiful’, it was ‘epic’. Such praise is often given with the best intentions, particularly in the belief that positive feedback, especially for children who don’t have much faith in themselves, might help to raise their self-esteem. But does it work?

Recent research by Eddie Brummelman and colleagues has tried to shed light on this question. In three studies, they looked at how adults dish out praise to children in both an experimental and naturalistic setting, and how children with varying levels of self-esteem take it. Their results suggest that overly positive praise might not have the intended effect for children who have low self-esteem.

In the first experiment Brummelman’s team asked a group of adults to read short descriptions of hypothetical children, described as either having high or low self-esteem. People were told about something that the child had done – say, solving a maths problem, or performing a song. After reading through the description, they were asked to write down any praise that they might give the child. Brummelman’s team found that about a quarter of the praise was overly positive (e.g. ‘that sounded magnificent!’), and that people were more likely to give more extremely positive praise to the children who had low self-esteem.

The researchers then tried to replicate these findings in a more naturalistic setting, by observing how parents interacted with their children (7–11 years old) when giving them a series of maths exercises at home. Brummelman and colleagues found a similar result to their laboratory experiment – about a quarter of the time, praise was overly inflated, and children who had lower self-esteem were given more inflated praise than those who had higher self-esteem.

In order to figure out whether this actually mattered or not, in the final experiment Brummelman’s team looked at how being given praise impacted on one particular aspect of children’s behaviour – challenge seeking. Two hundred and forty children first completed a questionnaire to assess their level of self-esteem, and then were asked to draw a copy of van Gogh’s Wild Roses. The children were told that a professional painter would then assess their drawing, and tell them what he thought of it. In reality, the painter didn’t exist, and children were simply given inflated praise, non-inflated praise, or no praise at all. Afterwards, the children were shown four complex and four easy pictures, and asked to have a go at reproducing some of them. Critically, they were told that if they picked the difficult picture, they might make a lot of mistakes, but they might also learn lots. In other words, the number of difficult pictures the children chose to draw was taken as a measure of challenge seeking.

Brummelman’s team found that if children with lower self-esteem had been given overly inflated praise, they were less inclined to seek a challenge in the second task – they would go for easy drawings over the harder ones, and therefore miss out on the chance for a new learning experience. On the other hand, children with high self-esteem were more likely to seek a challenge after being given inflated praise. Interestingly, the only difference between the inflated and non-inflated praise was a single word – incredibly (‘you made an incredibly beautiful drawing!’ versus ‘you made a beautiful drawing!’).

What the study doesn’t tell us is why children with low-esteem might avoid challenges in these circumstances. The authors suggest that inflated praise might set the bar very high for children in the future, and so inadvertently activates a self-protection mechanism in those with low self-esteem – although they acknowledge that they didn’t actually measure this in the study.

At any rate, the finding builds on a number of experiments conducted in recent years showing that positive praise isn’t necessarily good for all children in all circumstances. For children with low self-esteem, although we might feel the need to shower them in adulation, this might end up having precisely the opposite effect. Even words like incredible can end up having a huge unintended impact – so when you’re telling children they’ve done a great job, choose your words wisely.

- Post written by guest host Dr Pete Etchells, Lecturer in Psychology at Bath Spa University and Science Blog Co-ordinator for The Guardian 

Never the earner, always the bride

In Social Forces 

Married men who have a more traditional ‘breadwinner role’ at home tend to have more negative views on women in the workplace, according to a series of studies led by Sreedhari Desai.

Using data from US national surveys, the researchers found that men in more ‘traditional’ marriages, where the wife was not employed, showed some discomfort with a gender-mixed workplace, being more likely to disagree with statements such as ‘If a mother chooses to work, it doesn’t hurt the child’. A second dataset, from a 2002 survey, suggested that traditionalists were less likely to see their workplace as running smoothly when it had a higher composition of women.

Turning to experimental work, Desai’s team showed that compared with those in a dual-earning marriage, traditionally married undergraduate students rated recruitment literature intended to attract job applicants as less effective when it contained cues of high female involvement in the company, such as all-female (vs. all-male) recruiter names and an equal-opportunity reference. 

The next experiment found managers just as susceptible; traditionally married managers were less likely to recommend a fictional candidate for an MBA programme if they were a woman. Interestingly, dual earners as a group gave higher ratings to the female than the male applicant.

Returning to survey data, the researchers were able to gather data across two data points of the British Household Panel Survey. In 1991 304 men were surveyed prior to marriage, and then in 1993 following marriage, using the same scale as study one used on attitude to women in the workplace. These attitudes didn’t predict the marriage structure men ended up in, but the type of marriage did affect subsequent attitudes to women at work, with a traditional set-up leading to less sympathy for women being represented in the workplace. 

So why might marriage be shaping these attitudes? Status construction theory suggests that we tend to use our own social conditions to extrapolate how the world works. If every day you engage in work duties while your wife focuses on home life, not only are you incentivised to believe that this is a sensible division of labour, but increasingly it will seem true to you, as your differential experiences give you more work-related resources such as contacts, knowledge and influence. 

The researchers conclude that attention could be given to ‘the challenging psychological position that men in traditional marriages face when alternating between their two daily realities’, and find ways to illustrate to these people that their personal life decisions may be driving their workplace attitudes, possibly in an unconscious fashion.

- By Alex Fradera, for the Occupational Digest (see www.occdigest.org.uk

How does stress affect your public speaking?

In Biological Psychology 

Having to give a talk or a speech in front of a large group of people is one of the scarier things we might find ourselves having to do. Ideally we want to give a flawless, well-rehearsed delivery, and getting too stressed is often linked to becoming – literally – lost for words. But is there any actual evidence for this link?

Tony Buchanan and colleagues have recently investigated what sort of aspects of speech and language are affected in stressful versus non-stressful situations. They asked 91 people to participate in a social-stress test, in which they had five minutes to prepare a speech, and then immediately deliver it. In the stressful condition, they had to imagine that they had been accused of shoplifting, and had to prepare a defence that they would deliver to the ‘store manager’ (an experimenter). Immediately after the speech, they were given a difficult mental arithmetic task. In the non-stressful condition, people spent five minutes preparing a summary of a travel article, which they then had to read aloud to camera. Immediately after, they completed a much simpler arithmetic task.

Buchanan’s team measured levels of the stress hormone cortisol in samples of saliva taken before the test, plus 10 and 30 minutes afterwards. They also measured heart rate, the speed at which people spoke during their speech, as well as the number of pauses and the number of ‘nonfluencies’ – words like um, er and hmm.

The stressful speech condition seemed to increase stress levels – the measures of heart rate and cortisol levels showed an increase in this condition compared to the non-stressful situation. However, some of the speech variables that the researchers looked at didn’t seem to be affected in the way that you might expect. Regardless of the stress condition, the speed at which people talked during their speech didn’t differ. Strangely, the number of nonfluencies was higher for the non-stressful speech than in the stressful one. The only detrimental thing that stress seemed to have an effect on was pause time – as they progressed through their speech, people tended to stop increasingly more often in the stressful condition as opposed to the non-stressful condition.

Buchanan and his colleagues acknowledge the limitations of their study – as it’s correlational in nature, we can’t say for sure whether increases in cortisol levels cause a greater number of pauses in speech production, or whether noticing that you’re pausing more often in the task causes your cortisol levels to increase.

That being said, it seems like pause time is important, because it is thought to be an indication of lexical retrieval processes – if more thought is required for a certain part of a speech, or harder words need to be used, you’re more likely to stop for a moment before saying them. In stressful situations, these retrieval processes take a longer time, and so you’re more prone to pausing. So this study seems like an interesting step forward in understanding specifically how stress affects different aspects of speech production – you might even say it gives us pause for thought.

I Post written by guest host Dr Pete Etchells (Bath Spa University and The Guardian) 

The material in this section is taken from the Society’s Research Digest blog at www.researchdigest.org.uk/blog. We are currently recruiting a new Digest Editor, and in the meantime you will find a number of ‘guest hosts’ writing for us. Visit the blog for full coverage, the archive and more.

Subscribe by RSS or e-mail at 


Become a fan at 


Follow the Digest editor at 



BPS Members can discuss this article

Already a member? Or Create an account

Not a member? Find out about becoming a member or subscriber