Earth Notes: On the Project CHARM Conference (2013-02)

Updated 2024-05-20.
Social norming vs gamification: notes from the Project CHARM conference at the British Academy. #norming #gamification

CHARM Conference Notes 2013-02-27

Venue: British Academy, Carlton House Terrace, London.

Three separate studies in the CHARM project were described during the proceedings.

Alan Warde: Behaviour, norms and sustainable consumption

We're back to behavioural analysis. (Behavioural Moment)

AGW implies radical changes in rich world consumption patterns.

Consumption not just purchase. It is ineradicable, beneficial, environmentally harmful.

Assuming tech won't solve the problem. So do we consume less or differently?

Strategies for changing behaviour: economic, social pressure, regulation, education. Not resoundingly successful because limited understanding of consumption dynamics.

Strategies for individual behaviour change: info, exhortation, nudge. (People don't act in accordance with their ethical values.)

Behavioural economics relies on steering and conditioning, not rational actors.

Nudge assumes that automatic/habitual/unconscious behaviour most important.

"Thinking fast and slow": Daniel Kahneman

*** CHARM is about the making and changing of norms. (Ie, collective behaviour rather than individual actors.) ... social engineering

Several dilemmas of collective behaviour change...

CHARM is trying to find out which interventions work and on what scale.

Ruth Rettie / Tim Harries : The Home Energy Study (domestic consumption feedback)


CHARM 3 studies:

Two theoretical themes in CHARM:

Home Energy Study... Feedback on domestic electricity consumption. Data collected with current-clamp meter and aggregated on central Web site. Took initial 2 weeks of baseline consumption readings.

Compared control (no feedback), individual feedback, individual + social norms.

Hourly usage graph was salient even if vertical scale (kWh) was opaque.

3% (not significant) reduction with feedback: would need larger sample. No difference between individual and social norms.

People we more interested in emails than Web site.

Social norms group looked at graphs more.

More viewing of tips correlated with reduced consumption.

Qualitative analysis (including probing, home tours, etc)

Discourse: often money is conventional; an important one is waste.

*** People more motivated by frugality (avoiding waste) than thrift.

Counter discourses: personal prefs, upbringing, ID, social expectations, etc.

CHARM feedback made people more aware of waste and consumption.

Smart meters may benefit from email/Web feedback, and *hourly* trend data.

DECC: Adam Cooper's response. (Available online.)


Q: even people claiming that they were different to the norm (eg we're a big family), took notice of norms at least some of the time.

(Feedback about specifics such as number of laundry loads per week for individual vs norm is probably more actionable.)

Suggestion to provide comparators against groups not too different from individual's current position to avoid being overwhelmed. (Rettie suggests that not suggested by survey responses.)

Kavita Patel / Tim Harries: iGreen: Facebook app to encourage sustainable behaviours

Fun, easy to use, light-hearted to fit FB environment.

In FB allowed 3rd party apps.

FB has some green games, but few green apps.

Good for samples of elusive groups such as young men.

Accessible via mobile media.

Little existing research uses apps.

Aim to test social norms approach.

Give users same quiz 7 times over several weeks to see how answers changed.

Groups: control, individual feedback, individual and norms feedback.

Typical question: How many of your clothes did you re-wear before washing?

2800 downloaded app but only 52 completed all 7 quizzes (sales funnel)...

No evidence that either sort of feedback helped.

Did the games and asking the questions change behaviour?

Did repetition increase salience?

Not all users were driven to download it because of the theme. Prizes may broaden the appeal.

The item on bottled water encouraged the "wrong" behaviour, ie drinking more bottled water and increasing plastic use.

Parisa Eslambolchilar: bActive

Most people have smartphones, which are loaded with sensors (accelerometers, GPS, etc, etc) and data connectivity and easy/convenient user access. bActive makes use of these features.

Ruth Rettie / Tim Harries...

Did the smartphone app encourage participants to walk more?

The app was always on (to measure *all* walking), unlike some similar apps.

Project required (the all male cohort) to carry their phone in a pocket.

(Ran on HTC Desire Android.)

Same 3-way split for control / feedback as other two.

Social norms as before showed participant compared to average and to top 20%.

Feedback groups took 39% more steps than control. Attr by participants to more short walks (43%) walk instead of car/pubtr (35%).

Type of feedback did not seem to make a difference.

Walking is just a means to an end: no awareness. "Walking is not exercise!" Measuring walking was interesting to participants.

Nominal 10,000 steps/week target, but users set their own, and day by day, etc.

Calories bad/dispiriting metric because a lot of steps to burn many!

Once working is a goal, then the incidental walking gets appreciated.

Also participants try to make up for 'bad days' looking at the "peaks and troughs" and "dips and spikes".

Users enjoyed liked looking at the graphs aka "digital traces" of their lives.

Motivation seems to be avoiding being a lazy "couch potato".

Participants very competitive and talked about "winning" and "beating the average" and liked smilies.

On days they thought that they could win they did extra steps, walked less when they did not think they could win. People well below (norms) were dispirited and did not want to look at the screen.

Social feedback creates competition but no evidence of additional walking.

Graphs were key (cross-gender across CHARM).

Ben Savage, DfT Response:

(DfT uses behavioural research already, eg in the "Think!" campaign.)

Average UK commute to work 9 miles. Max typical walk to work that is feasible is 1 mile. Almost no one walks to work at/above 2 miles.

Women and men have significantly different travel patterns.

Panel point: do pilots!

Closing comments:

Multi-domain research has some real advantages/insights.

Test effects with same app showing different control groups diff behaviour.

Design (eg UI) bias is important but not necessarily well reported/studied.

People fascinated by their digital traces, and reflect on them.

Both individual and social feedback are useful.

There is no theory that combines individual and norms.

"Nudges" are there, but maybe theory weak.