Providing information promotes greater public support for potable recycled water

Providing information promotes greater public support for potable recycled water

Accepted Manuscript Providing information promotes greater public support for potable recycled water Kelly S. Fielding , Anne H. Roiko PII: S0043-135...

912KB Sizes 0 Downloads 37 Views

Accepted Manuscript Providing information promotes greater public support for potable recycled water Kelly S. Fielding , Anne H. Roiko PII:

S0043-1354(14)00347-9

DOI:

10.1016/j.watres.2014.05.002

Reference:

WR 10656

To appear in:

Water Research

Received Date: 10 February 2014 Revised Date:

30 April 2014

Accepted Date: 2 May 2014

Please cite this article as: Fielding, K.S, Roiko, A.H, Providing information promotes greater public support for potable recycled water, Water Research (2014), doi: 10.1016/j.watres.2014.05.002. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

ACCEPTED MANUSCRIPT 1 2

Providing information promotes greater public support for potable recycled water

3

5

Kelly S Fielding1

6

Anne H Roiko2

7

SC

1. Institute for Social Science Research, The University of Queensland, Brisbane Queensland, 4072, Australia, [email protected] 2. School of Medicine, Gold Coast Campus, Griffith University, Queensland, 4222, Australia, [email protected]

M AN U

8 9 10 11 12 13 14 15

EP

Corresponding author: Kelly Fielding, Institute for Social Science Research, The University of Queensland, Brisbane, QLD, 4072, [email protected], ph: +61 7 33468725

AC C

19

TE D

16 17 18

RI PT

4

2

ACCEPTED MANUSCRIPT 20

Abstract

21 22

In spite of the clear need to address water security through sourcing new and alternative

24

water supplies, there has been marked resistance from some communities to the introduction

25

of recycled water for potable use. The present studies tested the effectiveness of providing

26

relatively brief information about the recycled water process and the safety of recycled water

27

on cognitive, emotional and behavioral responses. Three information conditions (basic

28

information or basic information plus information about pollutants in the water, or

29

information that puts the risk of chemicals in the water in perspective) were compared to a no

30

information control condition. Across three experiments there was general support for the

31

hypothesis that providing information would result in more positive cognitive, emotional, and

32

behavioral responses to recycled water. Information increased comfort with potable recycled

33

water and, in general, participants in the information conditions expressed more positive

34

emotions (Experiment 1 & 3), less negative emotions (Experiment 3), more support

35

(Experiment 1 & 3), and lower risk perceptions (Experiment 1 & 3) than those in the no

36

information control condition. Participants who received information also drank more

37

recycled water than control participants (Experiment 1 & 2, although the differences between

38

conditions was not statistically significant) and were significantly more likely to vote in favor

39

of the introduction of a recycled water scheme (Experiment 3). There was evidence, however,

40

that providing information about the level of pollutants in recycled water may lead to

41

ambivalent responses.

AC C

EP

TE D

M AN U

SC

RI PT

23

42 43

Keywords: potable recycled water, public acceptance, information

3

ACCEPTED MANUSCRIPT 44

45

1. Introduction With an estimated 700 million people currently experiencing water stress or water scarcity (World Bank, 2010) and two out of three people predicted to be living in water

47

stressed areas by 2025 (United Nations Environmental Program, 2008), access to water is a

48

critical issue globally. In spite of the clear need to address water security through sourcing

49

new and alternative water supplies, there has been marked resistance from communities to

50

some alternative water resources. For example, proposed recycled water projects in the USA

51

and Australia have met with community resistance that, in some cases, has prevented their

52

successful implementation (Hurlimann & Dolnicar, 2010; Po, Kaercher, & Nancarrow,

53

2003). These cases include San Diego (USA) and in Australia, the Sunshine Coast,

54

Toowoomba, and the region of south East Queensland. These examples highlight the critical

55

role that community acceptance plays in uptake and implementation of alternative water

56

systems.

SC

M AN U

TE D

57

RI PT

46

In an attempt to understand community opposition, a growing body of research has focused on identifying the determinants of support for potable recycled water schemes (i.e.,

59

recycled water used for drinking water purposes). This research has shown that risk

60

perceptions, which are subjective assessments of the magnitude and likelihood of negative

61

outcomes associated with the risk (Dillon, 2000; Dolnicar & Hurlimann, 2009; Hurlimann,

62

2007; Marks, Martin, & Zadaroznyj, 2008; Nancarrow, Leviston, &Tucker, 2009), trust in the

63

agencies responsible for delivering water recycling schemes(Marks, 2004; Nancarrow et al.,

64

2009; Po et al., 2003) and perceived knowledge, which is what people think they know

65

(Dolnicar, Hurlimann, & Grun, 2011; Hurlimann, Hemphill, McKay, & Geursen, 2008;

66

Marks, Cromar, Fallowfield, & Oemcke, 2002) have been shown to predict acceptance of

67

recycled water schemes consistently. These studies have shown that there is greater

68

acceptance when risk perceptions are lower and trust and perceived knowledge is higher.

AC C

EP

58

4

ACCEPTED MANUSCRIPT 69

70

1.1 Past research on recycled water communication Only recently have researchers moved beyond the focus on identifying the predictors of recycled water acceptance to address the question of how to promote greater acceptance of

72

recycled water schemes. Providing information about the water recycling process seems an

73

intuitive and appropriate approach given that people report not being well-informed about

74

recycled water (Dolnicar & Hurlimann, 2009). To date, though, there are just a few studies

75

that have empirically tested the effectiveness of information provision. Simpson and

76

Stratton’s (2011) study of residents of south east Queensland, Australia found that

77

participants who were exposed to a 47 page online information booklet reported higher

78

perceived knowledge and acceptance of recycled water for drinking compared to those who

79

did not view the booklet. Dolnicar et al. (2010) found that Australians increased their

80

willingness to use recycled water for most purposes after receiving simple information about

81

the recycling process compared to before receiving the information. In another study of

82

Australians, Roseth (2008) pre-tested attitudes and willingness to use recycled water for

83

various purposes, provided an information pamphlet to half of the sample and no information

84

to the other half and then post-tested attitudes and willingness again. The information

85

pamphlet provided information about why water should be recycled, how it is recycled,

86

safety of the water for various purposes including cooking and drinking, how it has been used

87

in Australia, and how the use in Australia compares to other countries. The results

88

demonstrated that information provision was associated with small but significant increases

89

in the number of people willing to use recycled water for seven out of 13 purposes (including

90

cooking and drinking), and recycled water was associated with more positive and less

91

negative attributes after information provision.

92 93

AC C

EP

TE D

M AN U

SC

RI PT

71

The present study builds on this past research in four important ways. First, the present studies test the effectiveness of relatively brief information about recycled water.

5

ACCEPTED MANUSCRIPT Although it is ideal to provide comprehensive information about the recycling process as

95

Simpson and Stratton (2011) did, in practice mass media campaigns compete for the time and

96

attention of people and therefore need to be brief and targeted. Second, the present studies

97

include a measure of actual behavior by offering participants the opportunity to drink water

98

that contains recycled water. Previous research has tested the effect of information on

99

attitudes toward recycled water and intended behavior for various purposes (Dolnicar,

RI PT

94

Hurlimann, & Nghiem, 2010; Roseth, 2008; Simpson & Stratton, 2011) but not on actual

101

behavior. Third, consistent with the recommendations of Dolnicar et al. (2010),the present

102

studies go beyond past research by incorporating a range of dependent variables (e.g. risk

103

perceptions, positive and negative emotion, trust) that are known to be associated with

104

recycled water acceptance. This is a crucial extension to the past literature as it can help to

105

identify the key mechanisms through which information has its effects. Finally, the present

106

studies test the effectiveness of providing information that specifically targets health risk

107

concerns, one of the key concerns that people express about recycled water for potable use.

108

The present studies test the efficacy of three different experimental conditions that provided

109

information about the process and safety of recycled water compared to a no information

110

control condition.

111

1.2 Provision of information about recycled water

M AN U

TE D

EP

AC C

112

SC

100

In the present study the focus is on potable recycled water as people are most resistant

113

to recycled water for this use (Dillon, 2000; Marks et al., 2008; Marks, 2004). Moreover,

114

given that perception of health risks is one of the key predictors of recycled water acceptance,

115

it makes sense to provide information that targets risk perceptions. The current studies

116

communicated basic information about the recycling process as well as information about the

117

safety of the recycled water. Because health concerns stemming from pollutants in waste

118

water are a key concern for community members (Dolnicar & Schäfer, 2009; A. C.

6

ACCEPTED MANUSCRIPT 119

Hurlimann, 2007), we also tested whether providing additional information that addressed

120

this concern would increase the effectiveness of the basic information.

121

The basic premise of providing information to improve acceptance of new technologies is the information deficit model, a perspective that assumes that if people

123

understand more about a topic or issue they will feel more positively and have greater

124

acceptance of it (e.g., Miller, 2004; Ziman, 1991). Although there are a range of critiques of

125

the deficit model (Sturgis & Allum, 2004), the findings from past studies that tested the

126

effectiveness of providing information about recycled water are consistent with the idea that

127

increasing knowledge through the provision of information can increase acceptance of

128

recycled water for a variety of uses (Dolnicar et al., 2010; Roseth, 2008; Simpson & Stratton,

129

2011). Nevertheless, the increases in acceptance in these studies were relatively modest

130

suggesting that there is scope to augment the effectiveness of the information being provided.

131

One way to allay fears about contaminants in recycled water is to provide information

132

about the likelihood of their presence. For example, providing information that there are very

133

low levels of pollutants in recycled water should be reassuring to people who are concerned

134

about this issue. The literature on responses to probability information raises questions

135

though about whether this type of assurance will be effective. Past research has shown that

136

many people do not understand probabilities (Miller, 2004), experts and lay people process

137

probability information differently (Kraus, 1992; Slovic et al., 1995) and people can be

138

relatively insensitive to probability information when they feel strongly about an issue

139

(Slovic, Peters, Finucane, & MacGregor, 2005). Hence, although it makes intuitive sense, it is

140

important to test the efficacy of providing information about the level of pollutants in

141

recycled water.

AC C

EP

TE D

M AN U

SC

RI PT

122

7

ACCEPTED MANUSCRIPT 142

Given the difficulty that people can have with processing probability information, an alternative approach is to translate risk information into concrete terms that can be more

144

easily understood. One way to do this is to use analogies that relate to people’s everyday

145

lives. This strategy is often recommended for the communication of medical risk information

146

(Edwards, 2003) although the efficacy of the approach has not been empirically established

147

(Barilli et al., 2010). The value of analogies is that they can provide comparative risk

148

information and a more concrete way to represent risks. Bostrom (2008) demonstrated that

149

people use analogies and comparisons as a way to understand risks. In a similar vein, past

150

research also suggests the need to provide rich context information if people are to judge

151

unfamiliar risks that have a low probability (Kunreuther, Novemsky, & Kahneman, 2001).

152

Although research by Barilli et al.(2010) suggests that verbal analogies do not necessarily

153

change risk assessments, the effect of analogies has not been tested in relation to information

154

about recycled water and therefore this research investigates this approach.

155

1.3 Scope of the present studies

SC

M AN U

TE D

156

RI PT

143

In the present study we present three experiments that test the effectiveness of providing information for changing the participants' responses to recycled water for potable

158

use (measured via cognitions, emotions, and behaviors). Within each of the three

159

experiments, we compare the responses of those in a control group who do not receive

160

information about recycled water with the responses of those in groups that receive

161

information of differing complexity. Participants in the three information conditions receive

162

one of three different types of information: 1) basic information that communicates facts

163

about the recycling process and the safety of recycled water, 2) the same basic information

164

plus information about the level of pollutants in the recycled water after treatment (pollutant

165

information), or 3) the same basic information plus information that provides an analogy of

166

how much water you need to drink to get a 100 mg dose of a pharmaceutical. We call the

AC C

EP

157

8

ACCEPTED MANUSCRIPT latter condition the perspective condition as it helps to put information about the level of

168

chemicals in the water into perspective. Experiment 1 draws on a student sample and

169

compares a control condition to the basic information and pollutant information condition.

170

Experiment 2 also draws on a student sample, but includes all three experimental information

171

conditions and compares the amount of recycled water participants in these conditions

172

consume in comparison to a control condition. Experiment 3 uses a general population

173

sample and tests all three information conditions against a control. On the basis of theory and

174

past research we advance the following hypothesis:

175

H1. Participants who receive information about recycled water will have greater knowledge

176

and will demonstrate more positive cognitive, emotional and behavioral responses to recycled

177

water than participants in the control condition who do not receive information about

178

recycled water. Specifically, experimental participants should express greater support, greater

179

comfort, more positive emotions, less negative emotions, lower risk perceptions and greater

180

trust in authorities who may deliver a potable recycling scheme. They should also drink more

181

water than those who do not receive information, and be more likely to vote for the

182

introduction of a potable recycled water scheme.

SC

M AN U

TE D

EP

183

RI PT

167

We do not make any firm hypotheses about differences between experimental conditions. One possibility is that participants will respond most positively to information that uses

185

analogy as this helps to put the risk of pollutants in the water into perspective.

186

1.4 Local conditions impacting on the present study The experiments were conducted in

187

south east Queensland, a region that had recently experienced the worst drought on record.

188

The Queensland government implemented a range of measures to address the drought

189

including water efficiency programs (Walton & Hume, 2011), as well as major infrastructure

190

projects such as the Western Corridor Recycled Water Project which was intended to provide

AC C

184

9

ACCEPTED MANUSCRIPT recycled water to augment drinking water supplies. Despite completion of the infrastructure

192

for the Western Corridor Recycled Water Project, the implementation of the scheme was put

193

on hold in 2009 and the Queensland government decided to introduce the recycled water into

194

the dam only when dam levels fell below 40% (Apostolidis, Hertle, & Young, 2011). The

195

project had attracted negative media attention (Roberts, 2008) which may have affected

196

public support for the scheme.

197

2. Experiment 1

SC

198

RI PT

191

We first tested our hypotheses with a convenience sample of university students. The advantage of using this population was that the hypotheses could be tested in a low-cost,

200

controlled environment that allowed water taste-testing at the conclusion of the study.

201

2.1 Participants and experimental design

M AN U

199

Participants were 63 undergraduate students from a large Australian university who

203

took part in exchange for course credit. Participants were assigned randomly to one of three

204

experimental conditions: 1) a no information condition, 2) a basic information condition, 3) a

205

basic information plus pollutant information. The sample comprised 14 males and 49 females

206

with a mean age of 19.35 (SD = 2.86).

207

2.2 Study procedure

EP

AC C

208

TE D

202

Participants signed up for a study about public responses to alternative water sources

209

and completed the questionnaire online in a computer laboratoryafter reading an information

210

sheet that provided ethical assurances (e.g. confidentiality, right to withdraw, researcher

211

contact details). All 63 participants completed the same questionnaire with the following

212

sections: First, a series of questions about participants’ level of comfort with various

213

technologies (including recycled water) followed by a second series of questions about their

214

demographic attributes. Participants in the information conditions were then presented with

10

ACCEPTED MANUSCRIPT information about recycled water for potable use (see Section 2.2.1 for details of this

216

information). After reading the information participants answered a series of questions that

217

assessed their responses to recycled water for potable use (see Section 2.2.2 for details of

218

these measures). Note that participants in the no information control condition did not receive

219

any information and went straight on to the dependent variables measures after completing

220

the demographic questions.

221

2.2.1 Experimental information conditions

SC

222

RI PT

215

Following the series of demographic questions, those in the ‘no information’ condition (Control group) went straight on to complete the items related to the dependent

224

variables in the questionnaire. Those participants exposed to the basic information condition

225

were provided with information taken from a Queensland government website

226

(http://www.seqwater.com.au/water-supply/water-treatment/purified-recycled-water) and

227

from a booklet aimed at improving public understanding of recycled water (Simpson, 2008).

228

The basic information described how purified recycled water (the term that has commonly

229

been used to describe recycled wastewater in Queensland) is cleaned and purified to very

230

high standards using advanced technologies and that it meets strict water quality and health

231

standards. A graphic of the water recycling process was provided followed by information

232

that the water meets the water quality and health standards of State (Queensland Health),

233

Federal (National Health and Medical Research Council), and world (World Health

234

Organisation) health agencies. In addition, the information drew on a water quality star

235

system (Simpson, 2008) with drinking water defined as five-star water and recycled water

236

defined as six-star water which is pure enough to use for processes such as kidney dialysis.

237

Participants in the pollutant information condition were presented with the same information

238

as the basic information condition plus information that: “The pollutants in this type of water

239

are undetectable, less than one part per trillion”. Whilst this additional information targets an

AC C

EP

TE D

M AN U

223

11

ACCEPTED MANUSCRIPT issue that can be technically quite complex, it was designed to be a simple statement that

241

would reassure participants that any pollutants that might remain in the water following

242

advanced treatment would be in very small amounts, below detection limits.

243

2.2.2 Dependent variable measures

244

2.2.2.1 Comfort with recycled water. The first series of questions were adapted from the

245

‘‘Comfort with Technology Scale” (Bruce & Critchley, 2012). Participants were asked how

246

comfortable they were with 12 technologies including clean coal, internet, genetically

247

modified plants for food, nuclear power plants, etc (0 = Not at all comfortable, 10 = Very

248

comfortable, 11 = Unsure). Note that participants who gave an unsure response to the

249

comfort with recycled water question were excluded from the analyses of this variable. In

250

Experiment 1, 2 participants gave an unsure response at Time 1 and 1 participant at Time 2.

251

Embedded within these questions were items assessing comfort with alternative water

252

sources including: drinking water that contains recycled water, drinking water that contains

253

treated storm water, and drinking water that contains desalinated water. The recycled water

254

item was used in the present study to assess responses before and after the provision of

255

information. The other water-related items were not of interest in the present study but were

256

included for other purposes and were in keeping with the study’s broader purpose of

257

investigating responses to alternative water sources.

258

2.2.2.2. Demographics. Following the Comfort with Technology scale a series of

259

demographic questions were asked including sex, age, and the degree program that

260

participants were enrolled in.

261

2.2.2.3. Support for potable recycled water was measured with four questions that asked: 1)

262

What is your overall view of purified recycled water? (1 = very unfavourable, 7 = very

263

favourable); 2) I would support adding purified recycled water to the water supply in

AC C

EP

TE D

M AN U

SC

RI PT

240

12

ACCEPTED MANUSCRIPT Wivenhoe Dam; 3) I do not want purified recycled water to be mixed with drinking water

265

(reversed); 4) Given the choice, I would not drink water that contained purified recycled

266

water (reversed). Responses to questions 2 to 4 were made on 7-point scales (1 = strongly

267

disagree, 7 = Strongly agree). These items were taken from previous research about

268

community responses to recycled water (Nancarrow et al., 2009) and the four support items

269

were averaged and formed a reliable scale (α = .86).

270

2.2.2.4. Comfort with recycled water. The single item from the Comfort with Technology

271

scale was repeated for all participants after being asked the support items. This allowed us to

272

examine whether there were changes in this variable as a result of receiving information.

273

2.2.2.5. Emotional responses to recycled water were assessed by asking participants how

274

much they felt three positive emotions (encouraged, relaxed, happy) and three negative

275

emotions (disgusted, anxious, uncomfortable) when they think about purified recycled water

276

(1 = not at all, 7 =extremely). Principle components analysis suggested two factors with

277

eigenvalues greater than 1 that explained 80% of the variance—a positive emotion factor and

278

a negative emotion factor. Hence, the positive emotion items were averaged to form a

279

positive emotion scale and a negative emotion scale was computed from the average of the

280

negative emotion items (both with α = .87).

281

2.2.2.6. Perceived risk was measured with three items: 1) Drinking purified recycled water

282

will pose a health risk to me; 2) Drinking purified recycled will not lead to health problems in

283

the community (reversed), 3) People are exposed to so many risks everyday that the risk of

284

purified recycled water is too small to worry about (reversed) with responses on 7-point

285

scales (1= strongly disagree, 7 = strongly agree). The mean of the three items formed a scale

286

with reasonable reliability (α = .65; note that using the individual risk items in the analyses

287

did not change the pattern of results). Higher scores indicated greater risk perceptions.

AC C

EP

TE D

M AN U

SC

RI PT

264

13

ACCEPTED MANUSCRIPT 2.2.2.7 Trust in authorities to manage a recycled water scheme was assessed. Participants

289

were asked how much they trusted (1 = no trust at all, 3 = some trust, 5 = complete trust): 1) a

290

Queensland Government body to oversee and regulate the whole recycling scheme, 2) a

291

Queensland Government body to ensure that the recycled water meets with the water quality

292

standards, 3) Queensland Health Department to set safe drinking water standards, 4) scientists

293

to produce the best science for safe drinking water from a recycling scheme. The average of

294

the four items formed a reliable scale measuring trust in authorities to administer safe

295

recycled water (α = .85).

296

2.2.2.8Perceived knowledge was measured by asking: How much do you think you know

297

about recycled water for drinking water purposes (1 = nothing at all, 7 = a lot).

298

2.2.2.9. Behavior was measured through offering participants the opportunity to drink water

299

that contained purified recycled water. A plastic cup with 300ml of water was placed on the

300

desks that the participants were sitting at and they were told that they could drink as much or

301

as little as they liked and the amount consumed by each participant was recorded in

302

milliliters. Note that participants sat in cubicles and therefore were not aware of the behavior

303

of other participants.

304

2.3 Results and discussion

SC

M AN U

TE D

EP

AC C

305

RI PT

288

Means, standard deviations, F values and partial eta-squared values (representing the

306

size of the effect of the experimental manipulation) are presented in Table 1. Inspection of the

307

means indicates that participants reported relatively high levels of support for potable

308

recycled water with means above the mid-point of the scale and risk perceptions were

309

generally low with means below the mid-point. Positive emotions were above the mid-point

310

of the scale for participants in the information conditions, but below the mid-point for the

311

control condition. Negative emotions were relatively low across all conditions and trust was

14

ACCEPTED MANUSCRIPT moderate, sitting around the mid-point of the scale. The amount of water consumed across all

313

conditions was relatively low (ranging from 15 to just over 41 milliliters from a possible 300

314

ml).

315

2.3.1 Preliminary analyses

316

RI PT

312

As Table 1 shows, a one-way analysis of variance (ANOVA) on the comfort with recycled water question (measured prior to information provision for the experimental

318

conditions) revealed that there were no pre-existing differences on this variable across

319

experimental conditions. The efficacy of information provision was confirmed with

320

participants in the information provision conditions reporting higher levels of perceived

321

knowledge about recycled water for drinking water purposes than the participants in the no

322

information control conditions (See Table 1). A number of univariate outliers (i.e., cases with

323

extreme scores greater than 3 standard deviations above or below the mean) were identified

324

on the key dependent variables: there was one outlier on the variables of support, risk,

325

negative emotions and trust and four on the amount of water consumed. The data for these

326

cases were removed for the relevant analyses.

327

2.3.2 Analysis of comfort with recycled water

M AN U

TE D

EP

A 2 (Time 1, Time 2) x 3(control, basic information condition, basic information plus

AC C

328

SC

317

329

pollutant information) mixed ANOVA with repeated measures on Time was conducted to

330

examine whether information increased comfort with recycled water. The expected Time x

331

condition interaction emerged, F(2, 57) = 5.33, p = 008, partial eta-squared = .16. We

332

followed up this interaction by examining the simple effects of Time within experimental

333

condition and these showed that although there was a tendency for comfort with recycled

334

water to increase from Time 1 to Time 2 in the control condition, this difference was not

335

significant (t = 1.73, p = .089), whereas the increase in comfort in the experimental

15

ACCEPTED MANUSCRIPT information conditions was significant: basic information condition, t = 6.36, p<.001; basic

337

information plus pollutant information condition, t = 3.42, p<.001 (see Table 1 for means).

338

These findings show a higher level of comfort with drinking recycled water following

339

reading information about the recycling process and assurances of the safety of the water.

340

2.3.3 Analysis of cognitive and emotional responses

341

RI PT

336

As Table 1 showed, one-way ANOVAs revealed the expected significant effect of experimental condition on support, risk, positive emotions and trust but not negative

343

emotions. One potential explanation for the lack of significant differences on negative

344

emotion is floor effects: the mean responses on this variable were low across conditions and

345

there may have been little room for participants to express even less negative emotion.

346

Tukey’s post-hoc tests were used to follow up the significant effects of experimental

347

condition and these showed that participants in both information conditions expressed

348

significantly more positive emotions and greater trust in authorities to deliver safe recycled

349

water than those in the control condition. The pattern was somewhat different on risk

350

perceptions and support: there was greater support and lower perceived risk in the basic

351

information condition than the no information control condition. Interestingly, the condition

352

in which participants also received pollutant information did not differ from the control

353

condition on support and risk perceptions.

354

2.3.4 Analysis of behavioral responses

M AN U

TE D

EP

AC C

355

SC

342

A similar pattern emerged on the amount of water consumed by participants. Although

356

the differences across the conditions were not significantly different, there was a tendency for

357

participants in the basic information condition to drink more water that contained recycled

358

water than those in the control condition (p = .065) but in the basic information plus pollutant

359

information condition participants did not differ from the control condition. Inspection of the

16

ACCEPTED MANUSCRIPT 360

standard deviations for this variable indicates that there was high variability within each

361

condition in terms of the amount of water consumed.

362

Table 1 also shows that the size of the effect of information on the variables ranged from .07 to .26. Rules of thumb for the size of experimental effects suggest that a partial eta-

364

squared of 0.06 can be considered a medium effect size and 0.14 a large effect (Cohen, 1988).

365

Hence, the effect of information provision on the dependent variables can be considered

366

medium to large.

367

3. Experiment 2

SC

RI PT

363

We sought to replicate the effect of information on actual behavior and therefore

369

conducted a second experiment with a student sample and the amount of recycled water

370

consumed as the dependent variable. In this experiment we also included the perspective

371

condition in which participants were given basic information plus information that attempted

372

to put risks associated with chemicals in the water into perspective.

373

3. Method

374

3.1 Participants and experimental design

TE D

EP

375

M AN U

368

In total 141 students from a large Australian university were recruited in exchange for AUD $10.00. Participants were assigned randomly to one of four conditions: 1) a no

377

information condition, 2) a basic information condition, 3) a basic information plus pollutant

378

information condition, and 4) a basic information plus perspective information condition. The

379

sample comprised 43 males and 98 females with a mean age of 21.95 (SD = 3.44).

380

3.2 Procedure and dependent measures

381 382

AC C

376

As in Experiment 1, participants signed up for a study about public responses to alternative water sources and were presented with the materials online in the laboratory. The

17

ACCEPTED MANUSCRIPT information conditions were the same as in Experiment 1 with the addition of the perspective

384

condition. Participants in the perspective information condition were presented with the same

385

information as the basic information condition plus they were told that: A person drinking 2

386

litres of six-star water daily would need 138,000 years to consume 100mg of a

387

pharmaceutical product (Simpson, 2008). We further clarified that 100mg was less than half

388

the active dose of a standard headache tablet. After reading the information participants

389

completed a questionnaire that pertained to a broader research program about alternative

390

water sources. Only one item from the questionnaire was relevant to the current study which

391

was the question (used in Experiment 1) that asked how much participants thought they knew

392

about recycled water for drinking water purposes. The other questionnaire data are not

393

presented in this paper. As in Experiment 1, participants were offered the opportunity to drink

394

water that contained recycled water. The procedure and measurement of behavior was the

395

same as Experiment 1.

396

3.3 Results and discussion

SC

M AN U

TE D

397

RI PT

383

Inspection of the means in Table 2 shows that, consistent with Experiment 1, participants in the experimental conditions reported knowing significantly more about

399

recycled water than those in the control condition. Table 2 also shows that participants in the

400

conditions provided with information also drank more water than those in the no information

401

control condition. In fact, participants in the basic information condition drank almost double

402

the amount of water as those participants in the control condition. Nevertheless, the

403

differences between conditions were not significant, no doubt due to the high levels of

404

variability within each condition and the size of the effect was small (Cohen, 1988).

405

4. Experiment 3

AC C

EP

398

18

ACCEPTED MANUSCRIPT 406

In Experiment 3 we sought to replicate the findings of the first two studies with a sample from the general community. As the study was conducted online we could not offer

408

participants the opportunity to drinking recycled water. In lieu of this behavioral measure we

409

included a question about whether participants would vote for or against the introduction of

410

recycled water to their local drinking water supply. This question has high ecological validity

411

as a referendum of this type was conducted in Toowoomba, a town in the South West region

412

of Queensland, Australia (Hurlimann & Dolnicar, 2009).

413

4.1 Participants and recruitment

SC

Participants were recruited from a permission-based online social research panel to

M AN U

414

RI PT

407

take part in a study about public responses to alternative water sources. A total of 305

416

participants agreed to take part with 133 (44%) males and I70 (56%) females (two did not

417

indicate their gender). The mean age of the participants in the sample was 47.25 (SD = 17.19;

418

range = 18 to 86) and participants had lived in south east Queensland an average of 25.50

419

years (SD = 19.55 years). Sixty percent of participants had gross household annual income of

420

less than AUD $90,000; 30% had completed primary or secondary school, 27% had a trade or

421

diploma qualification and 43% had university undergraduate or postgraduate education.

422

4.2 Procedure and measures

EP

AC C

423

TE D

415

Experiment 3 followed the same procedure as Experiment 1, except that participants

424

were not brought into the laboratory. Participants were randomly allocated to one of four

425

conditions: 1) no information control, 2) basic information, 3) basic information plus

426

pollutant information, 4) basic information plus perspective information. These conditions

427

were operationalized in the same way as in Experiment 1 and 2. After reading an information

428

sheet that provided ethical assurances participants first completed the Comfort with

429

Technology scale that was used in Experiment 1 that included the single item asking about

19

ACCEPTED MANUSCRIPT 430

comfort with recycled water followed by demographic questions. Participants in the

431

information conditions were then presented with information about recycled water.

432

Following the provision of information (or straight after the demographic questions for participants in the no information condition), participants completed the same set of

434

questions that were used in Experiment 1 to assess perceived knowledge, support for potable

435

recycled water (α = .93), perceived risk (α = .90), trust (α = .93) and positive (α = .94) and

436

negative emotions (α = .92). Participants were again presented with the single item asking

437

about their comfort with recycled water after the support items in order to assess whether

438

there were changes in comfort as a result of the information provision (as in Experiment 1).

439

Participants who gave an unsure response to the comfort with recycled water question were

440

excluded from the analyses of this variable. In Experiment 3 18 gave unsure responses at

441

Time 1 and 11 at Time 2. As noted above, one additional variable was included in

442

Experiment 3: participants were asked how they would vote if a referendum was held about

443

the introduction of purified recycled water to their local water supply. Three response options

444

were offered: I would vote in favour of the introduction of purified recycled water to the

445

drinking water supply, I would vote against the introduction of purified recycled water to the

446

drinking water supply, I’m unsure how I would vote.

447

4.3 Results and discussion

SC

M AN U

TE D

EP

AC C

448

RI PT

433

Means, standard deviations, F values for the one-way ANOVAs and partial eta

449

squared values are presented in Table 3. Inspection of the means shows that participants did

450

not report high levels of knowledge about recycled water, however, they expressed moderate

451

levels of comfort with recycled water for drinking purposes with means above the mid-point

452

of the scale at Time 1 and Time 2. Support for recycled water was above the mid-point of the

453

scale in the information conditions and therefore moderate. Participants did not report high

20

ACCEPTED MANUSCRIPT levels of positive or negative emotions about recycled water, with all means below the mid-

455

point of the scale. The means also suggest that, on average, participants did not trust

456

authorities to manage a recycled water scheme.

457

4.3.1 Preliminary analyses

RI PT

454

As in Experiment 1, anANOVA on the comfort with recycled water question

459

(measured prior to information provision for the experimental conditions) revealed that there

460

were no pre-existing differences in comfort with recycled water across experimental

461

conditions. Note that regression analyses controlling for demographic variables showed that

462

the inclusion of these variables did not change the pattern of results reported below.

M AN U

463

SC

458

The one-way ANOVA revealed a significant effect of experimental condition on perceived knowledge. Tukey’s posthoc tests showed that participants in the information plus

465

pollutant information condition and information plus perspective information condition

466

reported higher levels of knowledge about recycled water than those in the no information

467

control condition. Participants in the basic information condition did not differ in their

468

perceived knowledge from the no information condition or the other information conditions.

469

4.3.2 Analysis of comfort with recycled water

EP

A 2(Time 1, Time 2) x 4(control, basic information condition, basic information

AC C

470

TE D

464

471

condition plus pollutant information, basic information plus perspective information) mixed

472

ANOVA with repeated measures on Time was conducted to examine whether the provision

473

of information influenced comfort with recycled water. The expected Time x condition

474

interaction emerged, F(3, 268) = 6.95, p<.001. Simple effects analyses examining the effect

475

of Time within experimental condition showed that participants’ comfort with recycled water

476

went down from Time 1 to Time 2 in the control condition, however, this difference was not

21

ACCEPTED MANUSCRIPT significant (t = 1.82, p = .07). In contrast, there was a significant increase in comfort in all of

478

the experimental conditions: basic information condition, t = 3.55, p<.001; basic information

479

plus pollutant information condition, t = 3.4, p<.001, basic information plus perspective

480

information, t = 2.83, p<.01 (see Table 3 for means).

481

4.3.3 Analysis of cognitive and emotional responses

482

RI PT

477

It is clear from inspecting the means in Table 3 that the provision of information

resulted in more positive responses to recycled water for potable use than when participants

484

did not receive this information. Participants who received information expressed more

485

support, more positive emotions, and less negative emotions. The latter effect differs from

486

Experiment 1 where information provision did not influence negative emotions. Note,

487

however, that the condition who received pollutant information did not differ in their negative

488

emotions from the control condition. Participants in the information conditions also perceived

489

less risk in relation to recycled water, although, again, for the condition that received

490

pollutant information, this difference was not significant (p = .075). Unlike Experiment 1,

491

the provision of information did not influence trust in authorities to manage a recycled water

492

scheme in that there were no significant differences across conditions. Using the rule of

493

thumb of 0.06 as a medium effect size (Cohen, 1988), the provision of information resulted in

494

medium or close to medium effects. The size of these effects is clearly smaller than in

495

Experiment 1.

496

4.3.4 Analysis of voting intentions

497

AC C

EP

TE D

M AN U

SC

483

When asked whether they would vote for or against the introduction of a potable

498

recycled water scheme, a chi-square analysis showed that the provision of information had a

499

significant effect on responses χ2 = 17.26, df = 6, 290, p= .008. Table 4 shows that

500

substantially more participants in the experimental conditions said they would vote in favour:

22

ACCEPTED MANUSCRIPT between 45% and 56% compared to 26% in the control condition. Indeed, in the condition in

502

which perspective information was provided, more than twice the number of participants said

503

they would vote in favor compared to the control condition.

504

5. General Discussion

RI PT

501

In the present study we tested whether providing information about the water

506

recycling process and the safety of recycled water for potable use could improve community

507

responses to this alternative water source. We also explored whether adding additional

508

information that addressed the safety of recycled water, namely, information about the level

509

of pollutants, or information that puts into perspective the amount of chemicals in the water

510

could increase the effectiveness of information provision. Across three experiments drawing

511

on student and community samples we showed general support for our hypothesis that the

512

provision of information about the safety of recycled water would be associated with greater

513

knowledge and more positive responses to recycled water for potable use. In this respect our

514

findings are consistent with previous research that has demonstrated the effectiveness of

515

information for increasing acceptance of recycled water (Dolnicar et al., 2010; Roseth, 2008;

516

Simpson & Stratton, 2011). None of the studies in the current paper showed superior effects

517

of adding additional information to address safety concerns and, in fact, the addition of

518

pollutant information may have decreased the effectiveness of the information (see below for

519

further discussion). The current studies went beyond previous research by examining the

520

effect of information on a range of cognitive, affective and behavioral responses to potable

521

recycled water. It is interesting to note that information most consistently influenced positive

522

emotional responses. Participants who received information about recycled water were more

523

comfortable with the idea of drinking recycled water after they received information than

524

before and they also felt more positive emotions than the participants who did not receive

525

information.

AC C

EP

TE D

M AN U

SC

505

23

ACCEPTED MANUSCRIPT 526

The pattern of results on risk and negative emotions was somewhat less clear cut. In general, support was higher and risk perceptions and negative emotions (in Experiment 3)

528

were lower when participants were provided with information, however, it appears that

529

augmenting basic information with information about level of pollutants in recycled water led

530

to similar responses on these variables to the control condition. Specifically, in Experiment 1,

531

participants who received pollutant information did not differ from participants in the control

532

condition on perceived risk or support and in Experiment 3 the pollutant information

533

condition did not differ from the control condition on perceived risk or negative emotions.

SC

One potential explanation for this pattern of findings is that the information about

M AN U

534

RI PT

527

pollutants focused participants on the notion of their presence in the water but did not put the

536

risk of these pollutants in perspective. Participants were told that “pollutants in this type of

537

water are undetectable, less than one part per trillion”. Lowenstein et al. (2001) document the

538

minor role that probability can have on emotions. They point out that events that have strong

539

positive or negative associations are largely uninfluenced by probability information. For

540

example, they note that the idea of winning the lottery elicits the same feelings whether the

541

chance of winning is 1 in 10,000 or 1 in 10,000,000. They conclude that the feelings

542

associated with risk-related decisions have “an all-or-none characteristic; they may be

543

sensitive to the possibility rather than the probability of negative consequences” (p. 276).

544

Although it may seem reassuring to tell people that the level of pollutants in the water is less

545

than one part per trillion, sensitivity to the possibility rather than the probability of pollutants

546

may elicit ambivalence in people’s responses to recycled water. They may respond positively

547

to the basic information that describes the recycling process and the safety of the water, but

548

they may also express their concern at the possibility of pollutants in the water. The social

549

psychological literature on attitude ambivalence, that is, attitudes that include both positive

550

and negative evaluations (Crano & Prislin, 2006), suggests that there are cognitive and

AC C

EP

TE D

535

24

ACCEPTED MANUSCRIPT behavioral consequences of attitude ambivalence including weaker associations between

552

attitudes and intentions and attitudes and behavior (e.g., Conner & Sparks, 2002). In the

553

context of responses to recycled water, people who have ambivalent attitudes could be more

554

mixed in their support for a recycled water scheme or more susceptible to information from

555

groups opposing these types of schemes. These findings highlight the importance of testing

556

the effectiveness of different types of information for communicating about recycled water

557

empirically. They also suggest the importance of further assessing the utility of providing

558

perspective information, which could help to negate or ameliorate the presence of

559

information about trace pollutants, information that authorities may feel compelled to

560

provide.

M AN U

SC

RI PT

551

In addition to extending past research by examining the influence of information on a

562

range of variables known to be associated with recycled water acceptance, the present studies

563

also examined whether cognitive and emotional responses were mirrored in people’s

564

behavioral responses. We did this by offering participants in Experiment 1 and 2 the

565

opportunity to drink recycled water and in Experiment 3 the chance to say whether they

566

would vote for or against the introduction of a potable recycled water scheme. To our

567

knowledge, previous research has not examined behavioral responses to recycled water

568

information. The lack of statistically significant differences in the amount of water consumed

569

across conditions means that we cannot draw strong conclusions about these findings. It was

570

clear that participants in the information conditions drank more water than those in the

571

control condition, with participants in the basic information condition drinking two to three

572

times what those in the control condition drank. It was also clear that there was a high level

573

of variability in the amount consumed within conditions. This may be due to factors external

574

to the experiment, for example, how thirsty participants felt, or to individual differences in

575

participant characteristics such as their worldviews (Kahan & Braman, 2006). One general

AC C

EP

TE D

561

25

ACCEPTED MANUSCRIPT 576

pattern that points to the potential influence of participant characteristics is that the effects of

577

information on the dependent variables were stronger in the student samples than the

578

community sample (as demonstrated by the partial eta squared values). On the measure of whether participants would vote for or against the introduction of a

RI PT

579

potable recycled water scheme in their area, a significant effect of information emerged with

581

participants from the general community who received information much more likely to vote

582

in favor. In comparison to the no information condition where only 26% of participants

583

would vote in favor of a recycled water scheme, participants’ likelihood of voting in favor of

584

a recycled water scheme rose by 19% higher in the basic information condition to 45% and

585

by 21% in the pollutant information condition to 47%. Those in the perspective information

586

condition who had received information that put the risks of the water in perspective were

587

more than twice as likely to vote in favor as those who did not receive information with 56%

588

voting in favor. Looking across the behavioral measures, our findings provide some tentative

589

evidence that providing information about the safety of recycled water can go beyond

590

influencing cognitive and affective responses and also lead to more positive behavioral

591

responses.

M AN U

TE D

EP

592

SC

580

One final anomalous finding is worth discussing. There were higher levels of trust in authorities responsible for managing a recycled water scheme in the information conditions

594

compared to the no information condition in Experiment 1 but not Experiment 3. Past

595

literature on risk has placed trust as a central mechanism in determining risk perceptions and

596

acceptance of risky technologies (Earle, Siegrist, & Gutscher, 2007; Lofstedt & Cvetkovich,

597

2008). The mixed findings for trust in our studies suggest that trust is either not so central to

598

participants’ judgments or not easily influenced by the provision of information that focuses

599

on process and safety. One factor to consider is that in our studies participants were not being

600

asked about a specific recycled water scheme and the authorities who would administer that

AC C

593

26

ACCEPTED MANUSCRIPT scheme. Rather, they were asked hypothetically about recycled water and the extent to which

602

they would trust a range of authorities (e.g. Queensland government, scientists) who may be

603

associated with administering this type of scheme. Trust judgments may become more central

604

or open to influence when a scheme is real rather than hypothetical and therefore, specific

605

authorities are responsible for administering the scheme (A. Hurlimann & Dolnicar, 2009; A.

606

Hurlimann et al., 2008; Ross, Fielding, & Louis, 2014).

607

5.1 Limitations

SC

608

RI PT

601

Key strengths of the present studies are the use of an experimental design that incorporated a no information control group and the inclusion of dependent variables that

610

tapped into cognitive (e.g. support, trust), emotional (e.g. positive and negative emotion) and

611

behavioral (water consumed, voting intentions) responses to recycled water. Despite these

612

strengths, a main limitation of the research is its generalizability to other regions and other

613

populations. The studies were all conducted in south east Queensland (SEQ), a region that

614

has experienced water stress in the last decade. The environmental context may play an

615

important role in how receptive people are to recycled water schemes and information

616

relating to them. Past research has shown that people have more pro-water conservation

617

attitudes and behavior in water scarce regions or during water scarce periods (Gilbertson,

618

Hurlimann, & Dolnicar, 2011; Trumbo, Markee, O'Keefe, & Park, 1999). Further research in

619

different regions of Australia and other parts of the developed world is therefore needed to

620

examine whether responses to recycled water information might similarly be affected by

621

environmental context. As discussed above, follow-up research should also examine whether

622

there are particular characteristics of community members that might influence reactions to

623

the information about recycled water. For example, in a qualitative study of Toowoomba

624

residents, supporters and opponents of recycled water differed in their political ideology and

625

worldviews and were motivated to process information in ways that supported their position

AC C

EP

TE D

M AN U

609

27

ACCEPTED MANUSCRIPT on the issue of recycled water (Price, Fielding, & Leviston, 2012). Research examining the

627

role of these variables in responses to recycled water information would help to build in-

628

depth knowledge of the most effective communication strategies that appeal to all sections of

629

the community. Finally, it may be useful for future research to test the effectiveness of

630

providing information that compares the quality of traditional potable water supplies with

631

potable recycled water supplies. This may be particularly important in efforts to build support

632

for direct potable reuse where an argument could be made that the recycled water is of much

633

better quality than ‘normal’ potable supplies (Khan, 2013).

634

5.2 Conclusions

SC M AN U

635

RI PT

626

As pressures on water resources become greater and there is a need to seek alternative sources to augment drinking water supplies, recycled water for potable use will be a critical

637

part of the mix of solutions. The findings of the present studies suggest that providing

638

relatively brief information about the recycled water process and the safety of recycled water

639

is an important step toward improving community responses to indirect potable recycling

640

schemes. The present research confirms that information influences a range of cognitive and

641

emotional dimensions related to recycled water acceptance and has the potential to also

642

influence behavioral responses. Our results were not conclusive about whether there is

643

additional benefit to adding specific information about the level of pollutants in the water

644

after treatment or information that puts the risks of the water in perspective, although on the

645

measure of whether people would vote in favor of a recycled water scheme, the provision of

646

perspective information seem to have the greater positive effect. Future research that

647

combines the pollutant information with the perspective information could illuminate whether

648

this combination may be optimal for producing more positive responses. Our findings that

649

information can increase positive responses to potable recycled water are consistent with the

650

practices adopted by agencies who have successfully introduced potable recycled water

AC C

EP

TE D

636

28

ACCEPTED MANUSCRIPT schemes: in places where intentional schemes are in place, there is evidence of concerted

652

communication and education efforts leading up to their introduction (Ross, Chapman,

653

Roiko, & Head, 2013). Nevertheless, communication efforts do not operate in a vacuum and

654

opponents can also influence the discourse around recycled water. Future research is needed

655

to identify the mechanisms that can optimize information transfer and diffuse opposition to

656

recycled water schemes.

AC C

EP

TE D

M AN U

SC

RI PT

651

29

ACCEPTED MANUSCRIPT Acknowledgements

658

This research was funded by Australian Research Council grant FT100100704.

AC C

EP

TE D

M AN U

SC

RI PT

657

ACCEPTED MANUSCRIPT 30

References

AC C

EP

TE D

M AN U

SC

RI PT

Apostolidis, N., Hertle, C., & Young, R. (2011). Water recycling in Australia. Water (Australia), 3, 869-881. doi: 10.3390/w3030869 Barilli, E., Savadori, L., Pighin, S., Bonalumi, S., Ferrari, A., Ferrari, M., & Cremonesi, L. (2010). From chance to choice: The use of verbal analogy in the communication of risk. Health, Risk & Society, 12(6), 546-559. Bostrom, A. (2008). Lead is like mercury: risk comparisons, analogies and mental models. Journal of Risk Research, 11(1-2), 99-117. Bruce, G., & Critchley, C. (2012). The Swinburne National Technology and Society Monitor. Melbourne: Swinburne University of Technology. Cohen, J. (1988). Statistical power analysis for the behvioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Conner, M., & Sparks, P. (2002). Ambivalence and attitudes. European Review of Social Psychology, 12, 37-70. Dillon, P. (2000). Water reuse in Australia: current status, projections and research. Paper presented at the Water Recycling Australia 2000, Adelaide. Dolnicar, S., & Hurlimann, A. (2009). Drinking water from alternative water sources: differences in beliefs, social norms and factors of perceived behavioural control across eight Australian locations. Water Science and Technology, 60(6), 1433-1444. Dolnicar, S., Hurlimann, A., & Grun, B. (2011). What affects public acceptance of recycled and desalinated water? . Water Research, 45(2), 933-943. Dolnicar, S., Hurlimann, A., & Nghiem, L. D. (2010). The effect of information on public acceptance - The case of water from alternative sources. Journal of Environmental Management, 91, 1288-1293. doi: 10.1016/j.jenvman.2010.02.003 Dolnicar, S., & Schäfer, A. (2009). Desalinated versus recycled water: Public perceptions and profiles of accepters. Journal of Environmental Management, 90(2), 888-900. doi: 10.1016/j.jenvman.2008.02.003 Earle, T. C., Siegrist, M., & Gutscher, H. (2007). Trust, risk perceptions and the TCC model of cooperation. In M. Siegrist, T. C. Earle & H. Gutscher (Eds.), Trust in risk management: Uncertainty and scepticism in the public mind. London: Earthscan. Edwards, A. (2003). Communicating risks through analogies. . British medical journal [online], 327, 749. http://www.bmj.com.ezproxy.library.uq.edu.au/content/327/7417/749 Gilbertson, M., Hurlimann, A., & Dolnicar, S. (2011). Does water context influence behaviour and attitudes to water conservation? . Australasian Journal of Environmental Management, 18(1), 47-60. Hurlimann, A., & Dolnicar, S. (2009). When public opposition defeats alternative water projects - The case of Toowoomba Australia. Water Research, 1-11. doi:10.1016/j.watres.2009.09.020 Hurlimann, A., & Dolnicar, S. (2010). When public opposition defeats alternative water projects - The case of Toowoomba Australia. Water Research, 44(1), 287-297. doi: 10.1016/j.watres.2009.09.020 Hurlimann, A., Hemphill, E., McKay, J., & Geursen, G. (2008). Establishing components of community satisfaction with recycled water use through a structural equation model. Journal of Environmental Management, 88, 1221-1232. Hurlimann, A. C. (2007). Is recycled water use risky? An Urban Australian community’s perspective. Environmentalist, 27, 83-94. doi: DOI 10.1007/s10669-007-9019-6

ACCEPTED MANUSCRIPT 31

AC C

EP

TE D

M AN U

SC

RI PT

Kahan, D. M., & Braman, D. (2006). Cultural cognition and public policy. Yale law and policy review, 24, 149-172. Khan, S. (2013). Drinking water through recycling. The benefits and costs of supplying direct to the distribution system. Melbourne, Victoria: Australian Academy of Technological Sciences and Engineering. Kraus, N., Malmfors, T., & Slovic, P. (1992). Intuitive toxicology: Expert and lay judgments of chemical risks. Risk Analysis, 12(2), 215-232. Kunreuther, H., Novemsky, N., & Kahneman, D. (2001). Making low probabilities useful. The Journal of Risk and Uncertainty, 23(2), 103-120. Lofstedt, R. E., & Cvetkovich, G. (2008). Risk management in post-trust societies. London: Earthscan. Lowenstein, G. F., Weber, E. U., Hsee, C. K., & Welch, N. (2001). Risk as feelings. Psychological Bulletin, 127(2), 267-286. Marks, J., Cromar, N., Fallowfield, H., & Oemcke, D. (2002). Community experience and perceptions of water reuse. Water Supply, 3(3), 9-16. Marks, J., Martin, B., & Zadaroznyj, M. (2008). How Australians order acceptance of recycled water: National baseline data. Journal of Sociology, 44(1), 83-99. doi: doi: 10.1177/1440783307085844 Marks, J. S. (2004). Advancing Community Acceptance of Reclaimed Water. Water 46-51. Miller, J. D. (2004). Public understanding of, and attitudes toward, scientific research: what we know and what we need to know. Public Understanding of Science, 13, 273-294. Nancarrow, B. E., Leviston, Z., & Tucker, D. I. (2009). Measuring the predictors of communities’ behavioural decisions for potable reuse of wastewater. Water, Science & Technology. doi: doi: 10.2166/wst.2009.759 Po, M., Kaercher, J. D., & Nancarrow, B. E. (2003). Literature review of factors influencing public perceptions of water reuse. In C. L. a. Water (Ed.), Technical Report 54/03. Price, J., Fielding, K., & Leviston, Z. (2012). Supporters and opponents of potable recycled water: Culture and cognition in the Toowoomba referendum. Society and Natural Resources, 25, 980-995. Roberts, G. (2008). Disease expert warns on recycled sewage. The Australian. Retrieved from http://www.theaustralian.com.au/archive/news/expert-warns-on-recycled-sewage/storye6frg6oo-1111117895549 website: Roseth, N. (2008). Community views on recycled water - the impact of information Research Report 48. Ross, V. L., Chapman, H. F., Roiko, A., & Head, B. (2013). Enhancing implementation and public acceptance of recycled water schemes: Lessons learnt from international case studies. Paper presented at the The 7th International Water Association Specialist Conference on Efficient Use and Managment of Water, Paris, France. Ross, V. L., Fielding, K. S., & Louis, W. R. (2014). Social trust, risk perceptions and public acceptance of recycled water: Testing a social-psychological model. Journal of Environmental Management, 137, 61-68. Simpson, J. (2008). From waste-d-water to pure water Retrieved from http://archive.nwc.gov.au/urban/more/from-waste-d-water-to-pure-water-reprint Simpson, J., & Stratton, H. (2011). Talking about water: words and images that enhance understanding Waterlines Report. Canberra: National Water Commission. Slovic, P., Malmfors, T., Krewski, D., Mertz, C. K., Neil, N., & Bartlett, S. (1995). Intuitive toxicology. II. Expert and lay judgments of chemical risks in Canada. Risk Analysis, 15(6), 661-675.

ACCEPTED MANUSCRIPT 32

AC C

EP

TE D

M AN U

SC

RI PT

Slovic, P., Peters, E., Finucane, M. L., & MacGregor, D. G. (2005). Affect, risk, and decision making. Health Psychology, 24(4), S35-S40. Sturgis, P., & Allum, N. (2004). Science in society: Re-evaluating the deficit model of public attitudes. Public Understanding of Science, 13, 55-74. Trumbo, C. W., Markee, N. L., O'Keefe, G. J., & Park, E. (1999). Antecedent precipitation as a methodological concern in attitude surveys on water conservation. Water Resources Research, 35(4), 1269-1273. Walton, A., & Hume, M. (2011). Creating positive habits in water conservation: the case of Queensland Water Commission and the Target 140 campaign. International Journal of Nonprofit and Voluntary Sector Marketing, 16, 215-224. doi: 10.1002/nvsm.421 Ziman, J. (1991). Public understanding of science. Science, Technology & Human Values, 16, 99-105.

ACCEPTED MANUSCRIPT

Table 1.Experiment 1: Means and standard deviations (in parentheses) for each condition, F values for the one-way ANOVAs and partial etasquared values No information

Basic information

Perceived knowledge

(N = 20) 2.76a (1.45)

(N = 21) 4.19b (1.40)

Basic Information + pollutant information (N = 21) 4.57b (1.17)

Time 1 comfort with recycled water

6.35a (3.10)

7.15a (2.52)

7.71a (2.67)

Time 2 comfort with recycled water

7.10a (3.16)

9.62b (1.56)

9.00b (1.84)

Support

5.23a (1.32)

6.33b (.99)

Perceived risk

3.38a (1.16)

2.59b (.89)

Positive emotions

2.95a (1.42)

4.32b (.98)

Negative emotions

2.15a (1.03)

Trust

3.28a (.72)

F value

Partial Eta Squared .26

1.25

.04

6.75**

.19

5.93ab (1.06)

4.96**

.14

3.28ab (.88)

4.02*

.12

4.05b (.94)

8.51***

.22

1.63a (.81)

2.06a (.73)

2.10

.07

3.84b (.56)

8.85***

.23

M AN U

TE D

EP 4.06b (.55)

SC

10.60***

AC C

RI PT

Dependent Variable

41.32a 25.71a 2.67+ .08 15.00a (14.24) (53.43) (27.26) Note. + p = .078; * p<.05; ***p<.001. Comfort with recycled water is measured on an 11-point scale ranging from 0 to 10. Water consumed is in millilitres. Trust is measured on a 5-point scale and all other dependent variables are measured on 7-point scales. Higher values represent more of the variable. Means within rows with different subscripts are significantly different from each other Water consumed

ACCEPTED MANUSCRIPT

Table 2.Experiment 2: Means and standard deviations (in parentheses) for each condition, F values for the one-way ANOVAs and Partial Eta Squared values Basic information

(N = 35) 3.09a (1.17)

(N = 35) 3.99b (1.17)

Basic Information + probability information (N = 35) 4.14b (1.35)

Basic information + perspective information (N = 36) 4.21b (1.42)

RI PT

Perceived knowledge

No information

SC

Dependent Variable

Partial Eta Squared

5.83***

.12

57.50a 107.43a 81.62a 93.33a 1.46 .03 (81.96) (121.92) (91.41) (110.27) Note. * p<.05; ***p<.001. Water consumed is in millilitres. Means within rows with different subscripts are significantly different from each other.

AC C

EP

TE D

M AN U

Water consumed

F value

ACCEPTED MANUSCRIPT

Table 3.Experiment 3: Means and standard deviations (in parentheses) for each condition, F values for the one-way ANOVAs, and Partial Eta Squared values

(N = 77) 3.95ab (1.51)

Time 1 comfort with recycled water

5.60a (3.32)

6.10a (3.36)

6.00a (3.37)

Time 2 comfort with recycled water

5.25a (3.52)

7.05b (3.39)

Support

3.59a (1.67)

4.52b (1.75)

Perceived risk

4.12a (1.54)

3.46b (1.56)

Positive emotions

2.57a (1.48)

3.26b (1.68)

Negative emotions

3.20a (1.54)

Trust

2.49b (1.54)

F value df (3,300)

Partial Eta Squared

3.79*

.04

6.60a (3.12)

1.13

.01

6.84b (3.77)

7.38b (3.36)

5.38***

.05

4.44b (1.75)

4.68b (1.67)

6.33***

.06

3.50ab (1.68)

3.17b (1.52)

5.00**

.05

3.27b (1.56)

3.44b (1.49)

4.74**

.05

2.63ab (1.59)

2.31b (1.47)

4.86**

.05

TE D

EP

Perceived Knowledge

Basic information + perspective information (N = 75) 4.14b (1.46)

RI PT

(N = 79) 3.44a (1.48)

Basic Information + pollutant information (N = 73) 4.12b (1.41)

SC

Basic information

M AN U

No information

AC C

Dependent Variable

2.79a 3.09a 3.09a 3.12a 1.49 .02 (1.11) (1.05) (1.26) (1.10) Note. **p<.01; ***p<.001. Comfort with recycled water is measured on an 11-point scale ranging from 0 to 10. All other dependent variables are measured on 7-point scales. Means within rows with different subscripts are significantly different from each other.

ACCEPTED MANUSCRIPT

Table 4. Experiment 3: Percentage of participants within each condition who would vote for or against the introduction of potable recycled water scheme in their local area

No information control

45.2

Basic information + pollutant information

47.2

M AN U

Basic information

56.3

AC C

EP

TE D

Basic information + perspective information

Vote against % 43.2

Unsure % 31.1

27.4

27.4

23.6

29.2

19.7

23.9

RI PT

Vote in favour % 25.7

SC

Experimental condition

ACCEPTED MANUSCRIPT Highlights:

EP

TE D

M AN U

SC

RI PT

Three experimental studies test the effectiveness of recycled water information Information focused on the recycling process and safety of the water Responses are assessed on cognitive, emotional and behavioral dimensions Information conditions responded more positively than the no information control

AC C

• • • •