Background
It is estimated that 2.5 or more billion people globally lack access to improved sanitation facilities [
1]. This sanitation deficit continues to leave the public exposed to a wide range of faecal contaminants responsible for a multitude of diseases, especially in densely populated slums [
2]. Estimates show that 4.2% or more of annual global mortality would be prevented if all people had access to safe drinking water, reliable sanitation and decent hygiene practices [
3,
4]. While some people lack total access to sanitation infrastructure, for others, it is a question of access to clean sanitation facilities. Using a dirty toilet exposes a user to the risk of contracting diseases, such as diarrhoea, and other intestinal and respiratory infections. The challenge of cleanliness is most prevalent in urban slums where several families share limited toilet facilities, for example more than 10 families sharing one toilet stance (room) [
5‐
7]. For cleanliness of the shared toilets to be guaranteed, it is imperative that user families are cooperative and collectively engage in their cleaning.
While there is substantial research around sanitation and its linkage to a wide range of preventable diseases [
3,
4], evidence on the cleaning behaviour of shared toilet users is still inadequate. More researchers and practitioners need to explore this area, which is fundamental to public and environmental health, especially in low income urban areas. We argue that performance of a behaviour, such as individual cleaning of a shared toilet, can be explained largely by psychosocial determinants and understanding the influence of social dilemma factors. The psychological determinants are itemized into five factor blocks in the RANAS model of behaviour change – one of the few models applicable to a wide range of water, sanitation and hygiene practices and interventions [
8]. The model synthesizes different social and health psychology theories and models and provides a structured approach for assessing, understanding, and explaining human behaviour as well as designing, implementing, and evaluating behavioural change-related interventions. The five conceptual block factors, each having a set of measurable variables include: risks, attitudes, norms, ability and self-regulating factors.
Risk factors [
9] relate to a person’s perceived vulnerability of contracting a disease, severity and consequences associated with the disease if contracted, and factual knowledge on disease exposure agents and how they can be prevented [
10].
Attitudinal factors indicate a person’s inclination to respond to a behaviour with some degree of liking or disliking the behaviour [
11]. Attitudinal factors can be categorized into instrumental, which are outcome expectancies (e.g. beliefs on costs in terms of money, time, effort and benefits associated with a desired behaviour and [
7,
12] and affective beliefs, which are feelings developed from thinking about a behaviour or its performance [
13‐
15].
Normative factors constitute descriptive norms, which reflect perceptions on behaviours typically performed by others, and injunctive norms, which show perceptions on behaviours typically approved or disapproved by people an individual considers important in their lives [
16].
Ability factors reflect a person’s confidence and belief to perform a behaviour [
16,
17]. Performance of a desired behaviour also needs a person to have traits of positive self-efficacy, which means abilities to organise and execute courses of action required to manage potential conditions, such as dealing with barriers that arise during the performance of the behaviour and recovery from setbacks [
18]. One major precondition of ability factors is action knowledge an assumption that one knows how to perform the desired behaviour [
19].
Self-regulating factors take precedence after the behaviour is in place and being performed but needs sustainability over time [
18,
20]. To consistently perform a desired behaviour, an individual should have the ability to manage conflicting goals and distracting situations [
21]. Self-regulating factors involve action control (strategy for a continuous standard evaluation of on-going desired behaviour) [
18], action planning (perceived thoughts on how to set up the behaviour and remembering and commitment to perform the desired behaviour [
22].
Each of the above RANAS model factors can be assessed using a structured questionnaire and may involve a set of variables for each factor [
8].
In contrast to the RANAS factors, social dilemmas are conflict situations characterised by decision-making processes, with most individuals making decisions that foster self-interests rather than those of groups to which they belong [
23,
24]. Yet, individuals would be better off making decisions that benefit the whole group [
25]. For instance, in the case of cleaning shared toilets, if all users of the shared toilet decided not to clean it, they would all receive lower payoffs, such as being exposed to the risk of diseases from the dirty toilet. Thus, the interest of integration of social dilemmas in this paper is on users of shared toilets’ cooperation, collective action, and commitment in their cleaning [
26]. As reported in some studies, proper hygiene practice is important to avert the risks of contracting diseases associated with unhygienic situations, such as using dirty toilets [
4,
27]. Sanitation research from the social dilemma perspective is still limited. Only a few studies were found that indirectly looked at the influence of some social dilemma factors, such as social norms on adoption of health behaviours [
28,
29].
In this study, we investigate the influence of social dilemma factors, such as group size, social identity, social motives, social norms, behaviour of others and communication on collective cleaning behaviour of shared toilet users.
First, we considered the size of the groups since this has been reported to have an influence on individuals’ cooperation in social dilemma situations. Studies have shown that the degree of cooperation declines with an increase in the size of the groups [
30,
31]. This argument is also evidenced in different sanitation studies that have shown the linkage between dirty shared toilets and the high number of user families [
6,
32].
Second, social identity is reported to positively influence cooperation among individuals. For example, in groups or in this case among users of a shared toilet to participate in cleaning, a sense of belonging or oneness as users of the toilet has a positive effect [
25,
33].
Third, social motive factors involve individual consideration of other people’s benefits while making individual decisions [
31]. Social motives among users of shared toilets could be manifested in their selfless cooperation in maintaining the cleanliness of shared toilets [
34,
35].
Fourth, social norms (shared beliefs and values that guide the way people behave or relate with each other) are reported in a number of studies as key in promotion of cooperation in resolving social dilemmas [
24,
36]. For example, social norms are important in the promotion of health behaviours, especially in the field of sanitation and hygiene [
28].
Furthermore, the behaviour of individuals as manifested in their decisions on whether to cooperate or not in social dilemma situations is influenced by their interpretations and observations of the behaviour of other persons in the same setting [
37]. Individuals are more likely to develop a cooperative behaviour if most of the others are cooperative [
38].
Lastly, communication has a cardinal influence in promoting cooperation and resolution of conflict situations, especially through face-to-face communication [
39,
40]. The importance of communication and using appropriate communication channels has also been of interest in sanitation and hygiene studies [
27,
41].
The objective of this study was to investigate the cleanliness situation of shared toilets in Kampala’s slums and the psychological and social dilemma factors influencing users’ collective cleaning behaviour and commitment. In regard to the operationalization of the RANAS model in understanding water and sanitation related-behaviours, studies have shown its effectiveness, such as in uptake of solar water disinfection (SODIS) [
15] and consumption of fluoride-free water [
42]. However, this is the first study applying RANAS and social dilemma factors to understand the cleaning behaviour of shared toilet users. While shared sanitation facilities take a broad spectrum of communal, public, and specific household shared facilities [
43], our study concentrates on the latter.
Results
The socio-demographic characteristics of the respondents are shown in Additional file
1. The majority of the respondents were female (75%), the mean age of the respondents was 31 years (range 18–75 years), and the majority interviewed were tenants (91.5%).
The mean number of people living in respondents’ households was about 4-persons (3.55) per household (range 1–30).
Cleanliness of shared toilets
Overall, over half of the shared toilets were reported clean (Table
1); however, interviewer observations showed that more shared toilets were very dirty than what was reported by the interviewees. There was a statistically significant Pearson correlation coefficient (P = 0.01) between interviewee perceived cleanliness and observed cleanliness by the interviewers.
Table 1
Perceived and observed cleanliness
Not dirty at all | 271 | 225 | 63.9 | 53.8 |
A little bit dirty | 44 | 41 | 10.4 | 9.8 |
Quite dirty | 13 | 22 | 3.1 | 5.3 |
Dirty | 65 | 59 | 15.3 | 14.1 |
Very dirty | 31 | 71 | 7.3 | 17.0 |
Total | 424 | 418 | 100.0 | 100.0 |
The reasons mentioned by respondents (n = 271) whose shared toilets were clean mainly related to the issue of cleaning them daily (62%) and cooperation (34.3%); other reasons (accounting for 3.7%) included every user household having a cleaning day, easy to clean toilet, few users, good toilet floor, and lockable toilet.
On the other hand, respondents (n = 153) whose toilets were dirty mainly attributed it to a large number of user families (40.9%) and lack of cooperation (30.2%); other reasons included bad use by some tenants (9.4%), misuse by children (5.4%), toilet almost full (3.4%), toilet full (2.7%), toilet having maggots (2%), not yet cleaned (2%), and misuse by outsiders (2%). Excreta on the walls and floor of the toilet room accounted for 2.1% of the respondents.
Cleaning of the shared toilets was largely attributable to gender. More than a third of the respondents (73.1%) reported that females were mainly responsible for the cleaning of shared toilets. About 15% of the respondents mentioned that males were mainly responsible for cleaning, and 9.9% of the respondents reported that both males and females were responsible for cleaning. Only 2.1% of the respondents mentioned that nobody was responsible for cleaning in their households.
The four main features reported by respondents for a clean toilet room were absence of excreta on the toilet floor (71.2%), no smell (64.2%), no flies (46%), and a dry toilet floor – not soaked with urine (41.3%). More information is shown in Table
2.
Table 2
Respondents’ understanding of a clean toilet and what is used in cleaning
Perceived understanding of a clean toilet
|
No faeces | 302 | 71.2 |
Toilet does not smell | 272 | 64.2 |
Toilet room has no flies | 195 | 46.0 |
Floor soaked with urine | 175 | 41.3 |
Faeces on toilet walls | 30 | 7.1 |
Toilet room has no maggots | 27 | 6.4 |
Toilet hole cover lid available | 20 | 4.7 |
Toilet ventilated | 5 | 1.2 |
Cleaning items
|
Water mixed with soap detergent | 313 | 73.8 |
Broom | 305 | 71.9 |
Plain water | 65 | 15.3 |
Cleaning brush | 46 | 10.8 |
Use a cleaning rag | 5 | 1.2 |
Smoking it using papers | 4 | .9 |
For cleaning frequency, 44.3% of the 424 respondents reported cleaning the shared toilet daily, 34.4% cleaned once or several times a week, 1.4% cleaned every second week, 5.4% cleaned once or several times a month, and 14.4% were not involved in cleaning at all. The respondents were using mostly brooms (71.9%) and a mixture of water with detergent (73.8%) to clean (Table
2).
The improved ventilated pit latrines were the most dominant type of toilet (74.8%), followed by simple pit-latrines (14.1%) and pour flush toilets (11.1%).
A number of diseases were reportedly associated with a dirty shared toilet. Out of 424 respondents, the diseases most frequently (multiple responses) mentioned were diarrhoea (70%), cholera (58.7%), candida (41%) and dysentery (17.2%).
Factors influencing shared toilet users cleaning behaviour
To determine the factors influencing collective cleaning of shared toilets by users, we assessed respondents’ self-reported cleaning frequency on the psychological (RANAS) and social dilemma factors using regression analysis.
RANAS and social dilemma factors
In the first step of the linear regression, RANAS variables accounted for 75.4% of the variation in respondents’ cleaning behaviour (Table
3). The introduction of social dilemma factors in the regression model increased the variance explained by the model to about 77%, as indicated by the R square = 0.77. There was no collinearity in the regressed variables (VIF below 6). The factors that were not statistically significant to respondents’ cleaning behaviour were excluded from the hierarchical linear regression. These included the affective factor to use a dirty toilet (RANAS), social identity factors of households relationships, behaviour of others, individuals’ cleaning cooperation and individuals participating less in cleaning, and unintended non-cleaning cooperation factor of individuals who were not held responsible for toilet dirt due to their inabilities (social dilemma).
Table 3
Linear hierarchical regression of respondent’s cleaning on RANAS and social dilemma variables
| Step 1 |
| (Constant) | .390 | .440 | | .886 | .38 |
Risk factors | Vulnerability to get disease | .052 | .080 | .017 | .656 | .51 |
Severity of disease | -.060 | .084 | -.019 | -.710 | .48 |
Attitude factors | Affective feeling - cleaning shared toilet | -.059 | .015 | -.126 | −3.998 | .00 |
Instrumental - cleaning time consuming | .071 | .047 | .055 | 1.511 | .13 |
Instrumental - cleaning effort | .039 | .040 | .035 | .976 | .33 |
Norm factors | Injunctive - approval to clean | .015 | .020 | .023 | .740 | .46 |
Injunctive - social pressure to clean | .017 | .026 | .018 | .657 | .51 |
Ability factors | Self-efficacy - cleaning difficulty | -.006 | .034 | -.007 | -.178 | .86 |
Self-efficacy - cleaning schedule | -.064 | .029 | -.063 | −2.239 | .03 |
Self-regulation factors | Action planning - cleaning daily routine | .505 | .048 | .521 | 10.538 | .00 |
Remembering to clean | .139 | .049 | .115 | 2.825 | .01 |
Cleaning commitment | .287 | .052 | .287 | 5.505 | .00 |
| Step 2 |
| (Constant) | .331 | .451 | | .735 | .46 |
Risk factors | Vulnerability to get disease | .031 | .079 | .010 | .398 | .69 |
Severity of disease | .023 | .084 | .007 | .269 | .79 |
Attitude factors | Affective feeling - cleaning shared toilet | -.060 | .015 | -.129 | −4.055 | .00 |
Instrumental - cleaning time consuming | .076 | .047 | .058 | 1.610 | .11 |
Instrumental - cleaning effortful | .048 | .040 | .043 | 1.205 | .23 |
Norm factors | Injunctive - approval to clean | .012 | .020 | .017 | .576 | .57 |
Injunctive - social pressure to clean | -.003 | .026 | -.004 | -.133 | .89 |
Ability factors | Self-efficacy - cleaning difficulty | -.026 | .036 | -.028 | -.713 | .48 |
Self-efficacy - cleaning schedule | -.069 | .029 | -.068 | −2.346 | .02 |
Self-regulation factors | Action planning - cleaning daily routine | .405 | .051 | .419 | 7.937 | .00 |
Remembering to clean | .118 | .049 | .097 | 2.410 | .02 |
Cleaning commitment | .237 | .053 | .237 | 4.462 | .00 |
Social motive factor | Respondents cleaning more than other users | .091 | .021 | .146 | 4.247 | .00 |
Communication factors | Talking frequency | .007 | .035 | .005 | .191 | .85 |
Talking ease | .030 | .033 | .032 | .903 | .37 |
Perceived efficacy factors | Shared toilet users’ cleaning cooperation | .042 | .036 | .043 | 1.169 | .24 |
Cleanliness confidence if other users are cooperative in cleaning | -.085 | .053 | -.045 | −1.601 | .11 |
Group dynamics factor | Cleaning team | .057 | .038 | .063 | 1.508 | .13 |
The negative, statistically significant, attitudinal affective factor associated with respondents’ cleaning of the shared toilet indicated that the more respondents dislike cleaning a shared toilet, the less they participated in cleaning. The negative, statistically significant, ability factor of a cleaning schedule indicated that respondents’ cleaning behaviour is less if their households have no cleaning roster regarding when to clean the shared toilet. On the other hand, the statistically significant self-regulating factors showed that respondents are more likely to frequently clean shared toilets if cleaning is part of their daily routine activities, it is easier to remember when to clean, and there is a cleaning commitment. Only one of the social dilemma variables was statistically significant. Respondents who believed they were cleaning more than the other shared toilet users participated more in collective cleaning, as shown by the social motive factor.
Respondents’ cleaning commitment
As shown in Table
4, social dilemma factors accounted for 67% (R Square = .67) of the variation in respondents’ collective cleaning commitment of the shared toilets.
Table 4
Linear regression of respondents cleaning commitment on social dilemma factors
| (Constant) | .726 | .259 | | .01 |
Social motives | Cleaning toilet more than other users | .166 | .016 | .338 | .00 |
Social identity | Shared toilet users’ relations | .086 | .020 | .219 | .00 |
Behaviour of others | Cleaning households | .003 | .001 | .072 | .02 |
Individual’s cooperation in cleaning | .018 | .039 | .013 | .65 |
Respondents cleaning less than other users | -.083 | .018 | -.162 | .00 |
Communication | Talking frequency with other users | .081 | .031 | .080 | .01 |
Easy to talk to other users | .154 | .028 | .212 | .00 |
Unintended non-cooperation | Individuals not held responsible | .049 | .043 | .032 | .26 |
Perceived efficacy | Shared toilet users’ cleaning cooperation | .036 | .033 | .048 | .28 |
Cleanliness confidence if other users are cooperative in cleaning | .136 | .046 | .092 | .01 |
Group dynamics | Cleaning team | .084 | .033 | .118 | .01 |
Social dilemma factors such as social motives, social identity, communication, and group dynamics, were positively related to respondents’ commitment to clean their shared toilets. Commitment was greater among respondents who believed they cleaned more than other users of the shared toilets, who positively related with other users, who easily talked with other users, and who felt they were part of a team with other users. However, while the perceived efficacy factor of household cooperation to clean shared toilets was not statistically significant, commitment was likely among respondents who had confidence that cleanliness of the shared toilets depended on the cooperation of all user households. Lastly, the behaviour of other households’ cooperation in cleaning of the shared toilets was not statistically significant. However, cleaning commitment by shared toilet users was less among respondents who reported cleaning less than the other toilet users.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
IKT is the primary author and provided conceptualization and study design, conducted the survey, data analysis and result interpretations, and manuscript writing. HJM provided academic supervisory support in conceptualization and study design, conducting the survey, data analysis and interpretation and manuscript review and revisions. All authors read and approved the final manuscript.