Centre for Confidence and Well-being

Skip to content

Research on the Effects of Positive Writing for S2 Pupils

The Centre for Confidence and Well-being carried out a series of positive writing exercises with S2 pupils in different schools across Scotland. The results were put together by The Centre and Dr Fabio Sani from Dundee University.

The Centre carried out a two week intervention in four schools across Scotland. Teachers were given written instructions to help facilitate one of three writing exercises with S2 pupils in their classrooms. The study found that all conditions, including the control group, had a positive effect on self-esteem, happiness, satisfaction with life, optimism and that depression scores decreased. Follow up interviews with pupils in the control group suggest that the exercise chosen was not a neutral activity which may be why pupils in this group reported positive changes in behaviour.

Further analysis suggested that there was a gender difference; boy’s happiness decreased after the positive writing intervention and their self-esteem stayed the same. Follow up interviews suggested that the boys found the exercise difficult. The results for the girls were more positive overall.

The content of the diaries, from the experimental group, were analysed; revealing 32 themes. The study showed that ‘doing sport’ and ‘small achievements’ were associated with higher levels of self-esteem within the whole sample. Males and females differed in what they wrote about on seven different themes. For example, females were more than twice as likely to refer to family compared to males. The interesting finding is that though males did not report family very often in their diaries, for those that did family was correlated with higher levels of well-being.

The aim of this study was to increase subjective well-being and decrease negative depressive symptoms in Scottish S2 pupils. This was to be achieved through various positive writing exercises.

Supporting literature
Research has found that a very simple task – keeping a nightly diary for one week listing three good things which have happened that day – can increase happiness and lead to a decrease in depressive symptoms (Seligman, Steen, Park & Peterson, 2005). Additional research (Emmons & McCullough, 2003) shows that counting one’s blessings every night for a period of 2 or 3 weeks increased physical well-being. Another study found that counting one’s blessings on a weekly basis for a period of 10 weeks increased positive emotion (Emmons & McCullough, 2003). Most recently Froh et al (2007) showed that adolescents who counted their blessings daily, for a period of two weeks, within their school setting displayed increased optimism, life satisfaction and gratitude; they also showed a decrease in negative affect. What’s more the research shows that the exercise need not be kept up indefinitely though it may be beneficial to undertake it about once a week ad infinitum.

Seligman believes that the reason why this exercise is so helpful in raising positive emotion and decreasing depression scores is because it trains the brain to pay attention to the positive. It is all too easy to pay particular attention to rejection, failure and a low mood, rather than what’s good in our lives (Baumeister et al, 2001). This could be an enjoyable conversation, a success, or something simple like enjoying a sunset.

The brain is hard-wired for negative emotion. It is the easiest thing for human beings to concentrate on as survival is dependant on us paying attention to what is wrong. This negativity bias was important for our ancestors who lived through times when life was dangerous for human beings. However, this negative bias is less important for us in contemporary life. What’s more, people are being encouraged to feel more negative as a result in changes in culture. For example, in past centuries people’s religion encouraged them to ‘count their blessings’ and feel grateful. The rise in marketing and consumer culture also encourages people to concentrate on what they don’t have or what’s not good about their lives.

The three good things exercise helps to train the brain to look for what’s good in our lives rather than to focus on what’s wrong or what we don’t like. By undertaking the exercise on a number of occasions we can train the brain to pay attention to the positive and to recall it, thereby increasing our experience of positive emotion. In short, this exercise is trying to get people into the habit of paying attention to the good things in their lives.

Study design
Pupils were asked to keep a diary for two weeks, and to write about a specific topic on a daily basis. They were allocated to three different conditions which were delivered by their class teacher:

Experimental condition (N=100) – Participants wrote about the “things that went well the day before”.
Dynamic condition (N=14) – Participants discussed about the “things that went well the day before” in a group setting, and then they individually wrote about that same topic in the diary. This activity was adapted to add in a social element; only one school participated in this exercise.
Control condition (N=85) - Participants wrote about the food they had consumed the day before.

The teachers were given written instructions for each of the interventions.
199 pupils (99 males and 100 females) all attending year 2, were selected from four schools across Scotland: school 1 (N=49), school 2 (N=53), school 3 (N=49), and school four (N=48). Consent was obtained from parents by the individual schools.

Participants completed a questionnaire including a set of measures. These were as follows:

Satisfaction With Life (SWL) scale. This is an instrument based on five statements concerning one’s level of satisfaction with life (e.g. ‘The conditions of my life are excellent’). Participants must specify how much they disagree or agree with each statement, using a seven-point scale ranging from 1 = ’strongly disagree’ to 7 = ‘strongly agree’.

Subjective Happiness Scale (SHS). This is an instrument including four items addressing the degree of the respondent’s happiness. Participants rate each item on a seven-point scale. For instance, the third item says: ‘Some people are generally very happy. They enjoy life regardless of what is going on, getting the most out of everything. To what extent does this description fit you?’  Participants respond on a scale ranging from 1 = ‘not at all’ to 7 = ‘a great deal’. 

Life Orientation Test-Revised (Lot-R). This instrument taps on one’s general orientation to life. It includes ten items, four of which are fillers. Therefore, the actual measure is based on six items (e.g. ‘I am always optimistic (positive) about my future’). Participants specify their level of disagreement or agreement with each item on a five-point scale ranging from 0 = ‘strongly disagree’ to 4 = ‘strongly agree’.

Rosenberg Self-Esteem (RSE) scale. This is an instrument assessing global self-esteem, including ten statements (e.g. ‘I feel that I have a number of good qualities’). Participants specify their degree of agreement or disagreement with each statement using a four-point scale ranging from 1 = ‘strongly agree’ to 4 = ‘strongly disagree’.

Center for Epidemiologic Studies Depression (CES-D) scale. This is an instrument measuring depression, based on twenty statements concerning how people have felt at different times (e.g. ‘I had crying spells’). Respondents express how often they have experienced the feelings reported in each items on a four-point scale ranging from 0 = ‘rarely’ to 3 = ‘most of the time’.

Following the two week intervention period the participants filled out the same measures again. The questionnaires and diaries were handed back from the teachers to the researchers to be analysed.

Table 1 and Table 2 report means and standard deviations in participants subjected to the three different interventions, in Phase 1 (before intervention) and in Phase 2 (after intervention) respectively.

These tables reveal that mean scores on self-esteem, satisfaction with life, subjective happiness, and optimism tend to be higher in Phase 2 than in Phase 1. Consistent with that, mean scores on depression tend to be lower in Phase 2 than in Phase 1. To assess whether differences between scores in Phase 1 and scores in Phase 2 were larger in a given condition than in others, we used a statistical technique called ANCOVA.  This revealed that different interventions produced the same effects on the dependent measures.

Table 1 - Descriptive statistics for participants subjected to intervention (before intervention)
condition   Mean Std. Dev.
Experimental            RSE 3.00 0.59
  SWL 4.95 1.16
  SHS 5.11 1.10
  CES-D 0.75 0.48
  Lot-R 2.35 0.87
dynamic RSE 3.11 0.40
  SWL 5.27 1.17
  SHS 5.16 0.90
  CES-D 0.69 0.37
  Lot-R 2.44 0.76
control RSE 2.88 0.54
  SWL 4.81 1.35
  SHS 4.95 1.14
  CES-D 0.83 0.47
  Lot-R 2.24 0.84
Table 2 - Descriptive statistics for participants subjected to intervention (after intervention)
condition   Mean Std. Dev.
experimental RSE 3.09 0.54
  SWL 5.11 1.17
  SHS 5.18 1.16
  CES-D 0.68 0.52
  Lot-R 2.47 0.84
dynamic RSE 3.21 0.52
  SWL 5.51 1.05
  SHS 5.37 1.04
  CES-D 0.52 0.29
  Lot-R 2.56 0.94
control RSE 2.97 0.58
  SWL 4.98 1.31
  SHS 5.11 1.09
  CES-D 0.67 0.46
  Lot-R 2.43 0.64
Analysis of gender scores
Despite all three conditions having an overall positive effect there were gender differences; boys became slightly less happy after the intervention compared to the control group. In addition to this the intervention had no effect on boy’s self-esteem. Overall the girls showed no negative effects.

Analysis of diaries
We analysed the content of the diaries filled in by participants in the experimental condition, in order to explore the nature of the themes that participants reported when writing about ‘things that went well the day before’. First, we searched for discrete themes reported in the entries of the diaries, and found 32 themes. Then, for each participant, we counted the number of entries in which each theme appeared. Because in each diary there were 42 entries (3 per each of the 14 days included in the intervention period), participants could score from 0 to 42 on each theme.

Table 3 lists the 32 themes that were reported in the diaries. This table also specifies the minimum and maximum number of times each theme was reported in a single diary, as well as the mean number of times each theme was used.

Table 3 – Descriptive statistics concerning themes reported in diaries
Themes Min Max Mean Std. Dev.
Appearance 0 4 0.33 0.77
Birthday 0 4 0.58 0.92
Chores 0 14 1.28 2.60
Cinema 0 2 0.19 0.42
Come home 0 4 0.17 0.58
Compliments 0 5 0.44 0.83
Computer (Using) 0 10 1.60 2.04
Cooking 0 6 0.92 1.38
Family 0 18 3.44 3.47
Food 0 11 2.36 2.22
Football 0 11 1.06 2.18
Help friend 0 6 0.42 1.00
Hobby 0 6 0.72 1.21
Holiday 0 2 0.28 0.47
Home improvement 0 4 0.23 0.67
Joking 0 9 1.39 2.22
Music 0 7 1.03 1.76
Negative humour 0 5 0.45 0.89
No homework 0 1 0.19 0.39
(Doing) Nothing 0 1 0.04 0.20
Not class 0 2 0.37 0.60
Pets 0 4 0.49 0.97
Sport 0 13 2.62 2.49
Reading 0 4 0.35 0.74
Classes 0 3 0.13 0.45
Shopping 0 11 1.70 2.12
Sleeping 0 4 0.40 0.80
Small achievement 0 8 1.55 1.88
Stuff 0 4 0.15 0.56
Talking 0 9 0.72 1.42
TV 0 15 1.60 2.08
Weather 0 4 0.35 0.69
We also looked at the correlations between the five measures (before intervention) and the 32 themes, to see whether well-being is associated with specific activities and situations reported in diaries. Concerning the sample as a whole, the only noticeable correlations concern self-esteem. This is positively correlated with both ‘small achievement’ (r = 0.32) and ‘sport’ (r = 0.32). Both correlations are statistically significant.
Analysis of diaries separately for male and female participants
Concerning gender differences in the reported themes, we performed a t-test for each theme, and found statistically significant differences on 7 themes. Table 4 includes a list of those themes upon which males and females differ significantly.  The minimum and maximum number of times each theme was reported in a single diary, and the mean number of times each theme was used, are also reported in this table.

Table 4 – Descriptive statistics concerning themes reported in diaries for males and for females
Gender   Min Max Mean Std. Dev.
Male (N=46) Appearance 0 2 0.09 0.35
  Computer (using) 0 10 2.26 2.52
  Family 0 8 2.30 2.32
  Football 0 11 2.02 2.84
  Joking 0 4 0.41 0.86
  Shopping 0 4 1.00 1.05
  Talking 0 4 0.28 0.86
Female (N=51) Appearance 0 4 0.55 0.97
  Computer (using) 0 6 1.00 1.25
  Family 0 18 4.47 4.00
  Football 0 2 0.20 0.53
  Joking 0 9 2.27 2.68
  Shopping 0 11 2.33 2.60
  Talking 0 9 1.12 1.69

This table shows that males refer to computers and football more often than females, while females refer to appearance, family, joking, shopping, and talking more often than males. Differences concerning family are particularly striking. On average, females refer to family almost twice as often as males. Also, note that in at least one case a female participant refers to family 18 times, while for males the top number of entries in which family is mentioned is 8.

At this point we looked at the inter-correlations between the five well-being measures (before intervention) and the 32 themes, separately for males and females. Some interesting patterns emerged.

Rosenberg Self-esteem (RSE) – This is positively correlated with both ‘sport’ (r = 0.40**) and ‘small achievement’ (r = 0.37*).
Satisfaction with Life (SWL) – This is negatively correlated with both ‘appearance’ (r = -0.30*) and ‘holiday’ (r = -0.32*), and it is positively correlated with both ‘football’ (r = 0.32*) and ‘talking (r = 0.31*).
Subjective Happiness Scale (SHS) – This is positively correlated with ‘family’ (r = 0.30*) and negatively correlated with ‘reading’ (r = -0.45**).
Center for Epidemiologic Studies Depression (CES-D) – This is negatively correlated with both ‘family’ (r = -.31*) and ‘not class’ (r = -0.37*).
Life Orientation Test-Revised (Lot-R) – This is negatively correlated with ‘computing’ (r = -0.29*) and positively correlated with both ‘family’ (r = 0.31*) and ‘not class’ (r = 0.34*).

Rosenberg Self-esteem (RSE) – This is positively correlated with both ‘doing nothing’ (r = 0.35*) and ‘reading’ (r = 0.33*).
Satisfaction withLife (SWL) – This measure is positively correlated with both ‘birthday’ (r = 0.29*) and ‘reading’ (r = 0.30*), and it is negatively correlated with ‘not in classroom’ (r = -0.39*).
Center for Epidemiologic Studies Depression (CES-D) – This is negatively correlated with both ‘reading’ (r = -0.35*) and ‘sleeping’ (r = -0.33*).
Life Orientation Test-Revised (Lot-R) – This is positively correlated with both ‘doing nothing’ (r = 0.29*) and with ‘reading (r = 0.36*).

In general, these results suggest that well-being in males is associated with a focus on sport, small achievements, holidays, talking, and family life, as well as with less time spent in the classroom and reading. Concerning females, well-being is associated with doing nothing, sleeping, reading, and class.

Interviews with the pupils
To find out more about the young people’s experience of the positive writing exercise, and also the food diary writing exercise, the Centre’s researcher went to two of the schools and talked to some of the young people from each of the groups.
Young people from the positive writing intervention group reported that it was very difficult to generate things that went well every time they were required to make a diary entry. Some reported that it was fairly easy to think of two things that went well, and that it was harder to generate ideas for more than two things. However, they also reported that over time this changed, at the end of the two weeks they were able to freely recall more things that went well. A number said ‘at the end I was finding it easier to write about good things’. This fits in with the theory that writing about good things trains the unconscious mind to attend to positive information which makes it easier to notice and recall things that go well in life. This finding also fits in with the overall results for happiness which increased after the two weeks. Though, as already been mentioned, the control group scores also changed.

Our researcher was aware of the gender difference which had been apparent both in the results and in the notebooks. Boys reported that they found it difficult to generate things that went well. One boy said: ‘I found it annoying because I wrote about two things and then I couldn’t think of anything else, I was then just saying something for the sake of it and it took the meaning away from the things I wrote about.’ Another boy said ‘I didn’t know what to write’. Some of the young people from the intervention group also reported that they would have found it helpful to have a discussion about what types of things they could write in the diary.
On talking to pupils in the control group our researcher found that some young people reported that they became more aware of what, and how much, they were eating: ‘I realise that I eat a lot’ and ‘I eat a lot of junk food’. These young people also reported that they made changes to their diet ‘I eat less now’ some said. Young people writing a food diary began to notice and pay attention to their behavior. They also reported making positive changes as a result, i.e. eating smaller amounts of food or more healthy food. From discussing the exercise with the young people, the overall consensus was that young people believed that food can make you happy. In other words, they believed that they were doing a positive intervention. One boy commented: ‘it’s like doing exercise you know it’s good for you’. Eating the right kind of food makes you feel good’.

All three conditions, including the control group, had a positive effect on the different measures: self-esteem, satisfaction with life, happiness, optimism and depression. Therefore it is unclear whether the experimental conditions or the control condition made the difference. Similar studies have used different control groups. For example Emmons & McCullough (2007) and Froh, Sefick & Emmons (2007) both used a daily hassles condition as a control. This involved participants recounting five hassles which occurred over the past day. Froh, Sefick & Emmons (2007) say that this is an appropriate control condition as it uses a different attention system. For example, focusing on hassles requires fault finding where as the counting blessings requires benefit finding. In the current study a follow up interview suggested that pupils in the control group became more mindful of what they were consuming; some pupils made positive changes to their behaviour. For these reasons the control group in the current study may not have been a neutral activity. Future research might explore this further. Other research may look into the gender differences. Similar to this study Felicia Huppert and colleagues found that elderly males showed some adverse effects after a positive writing intervention.

The following papers support the ideas behind this study and may be of interest.
Baumeister, R.F., Bratslavsky, E., Finkenauer, C., Vohs, K. D. (2001). Bad is stronger than good. Review of General Psychology, 4, 323-370.
Emmons, R.A., & McCullough, M.E. (2003). Counting blessings versus burdens: an experimental investigation of gratitude and subjective well-being in daily life. Journal of Personality and Social Psychology, 84, 377-389.
Emmons, R.A.(2007) Thanks: How the New Science of Gratitude Can Make you Happier. Houghton Mifflin Company.
Emmons, R. A., & McCulloch, M. E. (2004). The Psychology of Gratitude. Oxford University Press
Froh, J.J., Sefick, W.J., Emmons, R.A. (2007). Counting blessings in early adolescents: An experimental study of gratitude and subjective well-being. Journal of School Psychology
Seligman, M. E P., Steen, T., Park, N., & Peterson, C. (2005). Positive psychology progress: empirical validation of interventions. American Psychologist, 60(5), 410-421.
Sheldon, K. M., & Lyubomirsky, S. (2006a). How to increase and sustain positive emotion: The effects of expressing gratitude and visualizing best possible selves. The Journal of Positive Psychology, 1, 73-82.
For more information about this study contact contact@centreforconfidence.co.uk

Click on the document below to access the full Positive Writing Report.


 Positive Writing Report (249 KB)

Centre Events Previous Centre Events External Events Carol's Talks