ERIC EJ889225: Encouraging Electricity Savings in a University Residential Hall through a Combination of Feedback, Visual Prompts, and Incentives

2010·0.14 MB·English

by  ERIC

JOURNALOFAPPLIEDBEHAVIORANALYSIS 2010, 43, 327–331 NUMBER2 (SUMMER2010)
ENCOURAGING ELECTRICITY SAVINGS IN A UNIVERSITY
RESIDENTIAL HALL THROUGH A COMBINATION OF FEEDBACK,
VISUAL PROMPTS, AND INCENTIVES
MARTHINUS J. BEKKER, TANIA D. CUMMING, NIKOLA K. P. OSBORNE,
ANGELA M. BRUINING, JULIA I. MCCLEAN, AND LOUIS S. LELAND, JR.
UNIVERSITYOFOTAGO,DUNEDIN,NEWZEALAND
Thisexperimentinvestigatedthecombineduseofvisualprompts,dailyfeedback,andrewardsto
reduceelectricityconsumptioninauniversityresidentialhall.Aftera17-daybaselineperiod,the
experimentalinterventionwasintroducedintheinterventionhall,andnochangewasmadein
thecontrolhall.Energyusagedecreasedintheinterventionhall,butenergyusagedidnotchange
appreciably in the control hall. In the intervention hall, mean daytime and nighttime savings
were16.2%and10.7%,respectively,comparedtosavingsof3.8%(day)and6.5%(night)inthe
controlhall.
Key words: dormitories, electricity savings, feedback, incentives, master-metered residences,
visual prompts
_______________________________________________________________________________
University students who live in residential Incentives increase the likelihood of encourag-
halls typically have little or no incentive to ing proenvironmental behavior change (Boyce
moderate behaviors such as electricity usage, & Geller, 2001). For example, Slavin et al.
because the amount they pay is not directly (1981) encouraged residents in master-metered
influenced by how much they use (Natasha apartment blocks to save electricity by provid-
Austin, Resident, Salmond College, personal ingrebateswhenusagewaslowerthanpredicted
communication, September 10, 2007). This amounts, producing decreases between 1.7%
lack of consequences for residents’ excessive and 11.2%. Abrahamse et al. (2005) reported
electricity usage behavior can have financial that education about electricity conservation by
repercussions for the halls themselves, as well as itself increased knowledge but did not neces-
environmental repercussions (Abrahamse, Steg, sarily encourage change; experiments using
Vlek, & Rothengatter, 2005). There also is a positive reinforcers were the most effective.
response cost in saving electricity, if only the Feedback has been shown to play an
effort of turning off a switch when it is not important role in behavior change (Hayes &
needed. Slavin, Wodarski, and Blackburn Cone, 1981; Siero, Bakker, Dekker, & van den
(1981) described how master-metered apart- Burg, 1996). Petersen, Shunturov, Janda, Platt,
ments (apartments in which electricity is and Weinberger (2007) investigated feedback
included in rent) used 35% more electricity and reward techniques to reduce the electricity
than individually billed apartments. consumption of dormitory residents. They
Prompting, feedback, and incentives have compared real-time and weekly Web-based
been used to encourage electricity conservation. feedback of usage to residents in dormitories
in conjunction with an interdormitory electric-
ity conservation competition over a 2-week
We acknowledge the cooperation and assistance of the
period with a preceding baseline. Dormitories
staff and students of Salmond College and Cumberland
Collegewho werethe real starsofthis process. withongoingfeedbackreducedconsumptionby
AddresscorrespondencetoL.S.Leland,Jr,Department 55%, and those with weekly feedback saved
ofPsychology,UniversityofOtago,P.O.Box56,Dunedin
31%. Savings were for lighting and appliances
9054,NewZealand(e-mail:[email protected]).
doi:10.1901/jaba.2010.43-327 only; room and water heating were not
327
328 MARTHINUS J. BEKKER et al.
included. Prompts, feedback, and reinforcers (day and night) meters in each hall. All meter
are clearly important in encouraging desired readings in the control hall were done between
behavior change (Clayton & Helms, 2009). 10:30 a.m. and 11:15 a.m., and 36 of the 39
The present study used these techniques to readings in the intervention hall were done
encourage occupants of a residential hall to between 8:29 p.m. and 9:03 p.m.; the remain-
reduce their electricity use. ing three readings were within a 2-hr window.
Reading procedures included obtaining the
date, time, a day reading, and a night reading.
METHOD
Interobserver agreement was calculated by
Participants and Setting
having two meter readers conduct simultaneous
The participants in this study were the recordings on 57%and 48%of thecontrol and
occupants of two residential halls situated intervention readings, respectively, with 100%
within 2 km of each other at the University of agreement in both cases. An agreement was
Otago in Dunedin, New Zealand. The halls reached if each observer’s meter readings were
wererandomlyassignedtoeitheracontroloran within 2 kWh of each other.
intervention condition. The control hall had
326 residents, consisting mostly of 1st-year Procedure
university students aged 18 to 20 years (59% Consent and approval from the Otago
female and 41% male). The intervention hall University Ethics Committee were obtained,
had 190 mostly 1st-year residents (63% female then daily electricity readings started at both
and 37% male). Most rooms were single residential halls, which constituted the baseline.
occupancy. Heating and hot water costs were The experiment took place from August to
not included in the study because they were October (winter and spring in the southern
steamgenerated.Therewerenocoolingsystems hemisphere) with a temperature range (daily
present. maximum) between 5 and 12 uC. The
experimental materials were put up in the
Design, Response Measurement, and intervention hall during a Monday morning
Interrater Agreement while the majority of the residents were
This study was conducted as a control group attending classes. We put up nine large posters
design with a 3-week preceding baseline. The (29.7 cm by 42 cm) around the hall, calling
preceding baseline consisted of electricity read- residents’ attention to a savings thermometer.
ings at both control and intervention halls and The wording on the posters indicated that
allowed us to compensate for any intrinsic residents could get a reward for electricity
differences between electricity use in the two savings, which would be ‘‘good for the
halls.Thecontrolconditionwasidenticaltothe environment’’ and ‘‘good practice for flatting’’
baseline condition and consisted only of covert (shared independent living). We also put up
daily electricity meter readings. The 3-week smaller (10.5 cm by 14.8 cm) notices in 89
intervention used a combination of feedback, locations around the intervention hall, with
incentives, and education to encourage reduc- specific electricity savings tips for adjacent
tion of electricity use. The measured variable clothes driers, televisions, light switches, and
was the amount of electricity recorded as used computers. Both halls had comparable facilities
during each day and each night period, and the in this regard. We used an ‘‘electricity savings
independent variable was assignment to either thermometer’’ (in the intervention hall), which
control or intervention conditions. was hand-drawn on a whiteboard with colored
The dependent variable was the ongoing markers to deliver daily feedback and outline
record of the kilowatthours (kWhr) on the two theincentivescheme.Weplacedthewhiteboard
ELECTRICITY SAVINGS 329
directlyinsidethemainentrance.Itconsistedof would equate to 19,734.97 kWhr and
a thermometer shape with the various reward NZ$1,344.30. These amounts were calculated
levelsshownonascalealongtheside,indicating by subtracting the percentage change from
the proportional progress towards each reward baseline in the control hall from the percentage
combined with written descriptions of the change from baseline in the intervention hall,
rewards. The incentive value increased as the for both day and night, to derive an adjusted
residentssavedmoreelectricity.Thelowestlevel percentage change, which should account for
(free coffee for a week) cost $150 in New extraneous variables that affected both halls.
Zealand (NZ) dollars, and subsequent levels
The adjusted percentage change was multiplied
went up in NZ$50 intervals, with the most
by the daily baseline average in the intervention
expensive reward valued at NZ$350 (movie
hall to derive a kWhr-per-day savings, for both
night, ice cream plus pizza party).
day and night, which was then multiplied by
To update the savings thermometer each
the respective cost for day and night usage to
day, we calculated the electricity savings by
give a monetary savings per day. The per-day
subtracting that day’s usage from mean
kWhr saved and monetary savings was then
baseline (defined for both halls as the period
multiplied by the relevant time period to
preceding intervention in the intervention hall)
determine savings over the intervention period
usage for each hall. Then we subtracted the
as well as extrapolated savings.
percentage savings in the control hall from the
Figure 1 shows that the percentage of
percentage savings in the intervention hall.
electricity saved during intervention in the
Subtracting the savings in the control hall
intervention hall compared to its baseline was
controlled for any weather or university sched-
much higher during both the day (16.2%) and
ule-related factors. The differential savings for
night (10.7%) time than in the control hall,
the intervention hall were converted to a
whose residents saved a mean of 3.8% during
monetary value (kWhr saved multiplied by cost
the day and 6.53% at night during the same
per kWhr, adjusting for the specific day and
period compared to their baseline. Savings
night billing rates). Each weekday, cumulative
achieved were greater than in Slavin et al.
progresstowardtherewardswasupdatedonthe
(1981), where mean savings of 6.2% were
feedback thermometer (the monetary amounts
made. However, they were lower than in
were not given because the required monetary
Petersen et al. (2007), where feedback was
savings may have appeared too small to provide
combined with an interdormitory competition
incentive). The study ended after 3 weeks of
and savings of 31% to 55% made. These
intervention, and the reward that was reached
differences support the argument that more
was announced. We then distributed a social
immediate and frequent feedback is more likely
validity questionnaire to give participants an
to produce behavior changes (Siero et al.,
opportunity to comment on the study and for
1996). Slavin et al. provided fortnightly
the researchers to assess opinions related to the
feedback, in contrast to the constant real-time
study.
feedback used by Petersen et al.
Of the 190 residents, 75 completed a social
RESULTS AND DISCUSSION
validity questionnaire regarding the interven-
Over the 3-week intervention period, the tion. The majority of residents stated they
interventionhallsaved3,700.31kWhrequating wouldcontinuetosaveelectricityintheabsence
to NZ$251.31 (adjusted for control saving). ofrewards(74%,usingachi-squaregoodnessof
This entitled residents to a movie night with fit test, x2 5 13.34, p , .01) and that the
snacks. Over a 16-week semester, these savings scheme should be continued in future years
330 MARTHINUS J. BEKKER et al.
Figure 1. Percentage savings relative tobaseline for both hallsfor daily andnightlyelectricity usage.
(89%, x2 5 43.9, p , .01). The mean effort 1 kWhr of electricity. This extrapolation
residentsputintosavingelectricity(ona1to5 assumes a continuous rate of savings equal to
scale,15noneatall,55alot)was3.1(SD5 that of the 3 weeks of this study.
1.05), showing that the individual residents There are additional unmeasured benefits to
perceived that they had not put in excessive the intervention, such as reduction of carbon
effort. emissions and reduced environmental impact
Continuing this scheme over a full academic (Petrie et al., 2007). Further, reduced usage of
year (with one semester of baseline and one of appliances reduces their general wear and tear,
intervention)ispredictedtobecosteffectivefor thus decreasing the need to maintain and
halls.Table 1summarizesthepredictedcostsof replaceexpensiveitems.Onamoregloballevel,
maintaining the intervention. If we were to decreased electricity usage lowers pressure on
subtract the costs in Table 1 from the savings electricity providers to build new plants, again
(NZ$1,344.30 minus $216.27) and split the reducing environmental impact. For the resi-
remainder evenly between rewarding the resi- dents, there were social benefits, such as
dents and savings for the hall budget, that net fostering a sense of positive social identity by
savingwouldbe$564.02.Basedonthisanalysis working together towards a common reward
there would be an estimated NZ$5.22 net (Vaughan & Hogg, 2002). There are also
return for every NZ$1 spent on setup and potential ongoing benefits for the residents,
maintenance of the intervention. From another whomayusetheelectricity-savingtechniquesto
perspective, it only costs NZ$0.01 to save save money once they leave the hall.
ELECTRICITY SAVINGS 331
Table 1
Costsof Maintaining the Intervention Based on1 Academic Year
Item Costperunit(NZ$) Numberneeded Totalcost
Posterprinting Largecolorposters $0.90 9 $8.10
Smallpostersoncoloredpapera $0.22for4 80 $4.40
Small(A6)postersonwhitepaper $0.07for4 16 $0.28
Labor Posterdesign $11.25perhr(NZ 1hr $11.25
minimumwage)
Puttingupposters 0.5hr $5.63
Baselinemeterreadingsandcalculations 0.5hr $5.63
Meterreadingsandcalculationsin 25minperweekfor16weeks $75.00
intervention(weekdays) (6.66hr)
Updatingthermometerandchecking 25minperweekfor16weeks $75.00
andreplacingposters(weekdays) (6.66hr)
Stationery Whiteboard $135.00 1(over5years) $27.00
Whiteboardmarker $2.99 1(over2years) $1.50
Permanentmarker $2.99for3 3(over2years) $1.50
Adhesivetapeforposters $1.99for2 1 $1.00
Total $216.27
aThis includes 7 replacement posters that were used after some small posters were damaged or removed. No large
posters required replacement.
This experiment demonstrated that the Clayton, M. C., & Helms, B. P. (2009). Increasing seat
belt use on a college campus: An evaluation of two
combination of visual prompts, feedback, and
prompting procedures. Journal of Applied Behavior
incentives can effectively encourage behavior Analysis,42,161–164.
change regarding electricity usage. The inter- Hayes, S. C., & Cone, J. D. (1981). Reduction of
residential consumption ofelectricity through simple
vention hall saved more than the control hall,
monthly feedback. Journal of Applied Behavior
enough to earn a reward for their residents. It
Analysis,14,81–88.
should be noted that the baseline and interven- Petersen, J. E., Shunturov, V., Janda, K., Platt, G., &
Weinberger, K. (2007). Dormitory residents reduce
tion periods were relatively short, making it
electricity consumption when exposed to real-time
more difficult to conclude that the trends seen
visualfeedbackandincentives.InternationalJournalof
would continue over long periods of time. In Sustainability inHigher Education,8,16–33.
addition, this experiment did not examine Petrie,S.,Wear,S.,Cameron,C.,Gulliver,S.,Leslie,K.,
& Tsui, K. (2007). New Zealand’s greenhouse gas
heating or hot water usage, because these were
inventory 1990–2005: The national inventory report
steam generated. Future studies could validate and common reporting format. Wellington, New
the present findings by investigating a larger Zealand: Ministry for theEnvironment.
Siero, F. W., Bakker, A. B., Dekker, G. B., & van den
sample size over a longer time period and may
Burg, M.T.(1996).Changingorganizational energy
include some measure of reducing hot water consumption behaviour through comparative feed-
and heating consumption as well. back. Journal of Environmental Psychology, 16,
235–246.
Slavin,R.,Wodarski,J.,&Blackburn,B.(1981).Agroup
REFERENCES contingency for electricity conservation in master-
metered apartments. Journal of Applied Behavior
Abrahamse, W., Steg, L., Vlek, C., & Rothengatter, T. Analysis,14,375–363.
(2005). A review of intervention studies aimed at Vaughan, G., & Hogg, M. (2002). Introduction to social
household energy conservation. Journal of Environ- psychology (3rd ed.). Sydney, Australia: Pearson
mentalPsychology,25,273–291. Education.
Boyce, T., & Geller, E. (2001). Encouraging college
students to support pro-environment behavior: Ef- Received April 4,2008
fects of direct versus indirect rewards. Environment Final acceptanceDecember 19,2008
andBehavior,33,107–125. Action Editor,Kenneth Silverman