Results Of Questionnaire And Analysis

Published Date: 02 Nov 2017

4.1 Introduction

The research strategy for this thesis informed two forms of data collection, questionnaires and interviews. Questionnaires were carried out using survey monkey, the data was then exported into Excel and SPSS for data analysis. Interviews were analysed in a qualitative manor in order to establish themes associated with the studied literature. The following sub-sections will present the findings. The Graphs accompanying the descriptive analysis can be found in Appendix L.

The questionnaire research aims to fulfil objective two, building upon the review of energy consumption in the domestic environment. Elements of smart home design have also been incorporated, promoting responses which may inform objectives three and four. Finally the use of interviews focused primary on smart monitors, to fulfil objective three of this thesis.

4.2 Results of questionnaire and Analysis

4.2.1 Descriptive Statistical analysis

Energy Conservation

Question 5 related to basic energy conservation opinions. Analysis of the raw data determined the central tendency and standard deviation. In total 90% of responses either 'Agree' to 'Strongly Agree' that they are concerned with energy conservation. The central tendency has been identified as 'Agree' from the mean, suggesting the majority of the sample are concerned about energy conservation. The standard deviation is low, suggesting variance around the mean is small.

Question 6 probed the motivation of users for reducing energy consumption. In total 74% of respondents suggested that money was the primary motivation, followed by 20% claiming it was environmental impact. Only 2% and 4% of respondents claimed it was either a social obligation or that they weren't motivated at all.

Smart Metering

Question 8 related to the installation of smart meters. From the mean, a central tendency of 'Agree' was established. The standard deviation of data at 0.971 suggests there is some deviation, thus some people sway to 'Unsure' as well as 'Strongly Agree'.

Question 9 explored recording of consumption information. The central tendency from the mean was established to be 'Unsure'. The standard deviation of 1.169, suggests the result is fairly inconclusive.

Question 10 related to the use of a dynamic tariff. The mean lays between 'Agree' and 'Unsure'. The standard deviation is 1.087, suggesting opinion is split.

Smart Home Design

Question 11 related to mediums of energy feedback. Of the sample 62% responded that 'Personal Computer/Laptop Based Feedback' would be most suitable. Followed by 22% saying 'Smart Phone Based Feedback'. Therefore, 'Personal Computer/Laptop Based Feedback' is most preferred.

Question 12 explored home automation, in the context of reducing user control. The mean of 2.34 determined the central tendency of users was 'Positive' towards saving energy at the loss of user control. The standard deviation of 1.038 suggests opinion is split.

Question 13 related to energy consumption information. From the mean of 1.91 a central tendency that users 'Agreed' information would be useful was established. Supported by 84% of respondents 'Strongly agree' to 'Agree'.

Question 14 asked the importance of appliance efficiency. The mean of 1.94 identified a central tendency of 'Agree'. This suggests that energy users are concerned about appliance efficiency. In total, 84% of respondents 'Strongly agree' to 'Agree' that appliance efficiency is important.

Question 15 related to comparing energy consumption. The mean of 2.55 determined a central tendency of 'Unsure'. The standard deviation of 1.04, suggests a result is inconclusive as to whether users favour or dislike the comparison of consumption information. Although the data suggest 54% of users 'Strongly Agree' to 'Agree', whereas only 19% of users 'Strongly Disagree' to 'Agree'.

4.2.2 Energy Conservation Hypothesis

A) Home Owners are more concerned about energy conservation than Rented and Student Rented Accommodation

The multiple comparison Tukey test (Table 11), shows a significance level of higher than 0.05, when comparing the Student Accommodation group against the Rented and Home Owner groups. Figure 6 reinforces this, showing a visual difference between them. To conclude, there is only a significant difference between Rented and Home Owners against the Student Accommodation group. This suggests the Student group was least concerned about energy conservation.

B) Young adults are generally less concerned about energy conservation

The Tukey test (Table 14) was carried out, following our ANOVA (Table 13), in order to determine if difference is present in all or some of the groups. Figure 7 shows a significant result, due to being less than 0.05 significance. The main hypothesis of "Young adults are generally less concerned about energy conservation" has therefore been proven, as the mean of middle aged and older adults favours energy conservation, more than young adults. It should be noted however, a larger sample and spread of age would yield more conclusive results.

Mean Housing

Table 10 - Housing ANOVA

Table 9 - Housing Descriptive

Advanced Statistical Analysis (A)

Figure 6

Table 11 - Housing Tukey Test

Mean Age

Table 13 - Age ANOVA

Table 12 - Age Descriptive

Advanced Statistical Analysis (B)

Figure 7

Table 14 - Age Tukey Test

4.2.4 Smart Metering Technology Hypothesis

A) Users who agree to have smart meters installed, should also support dynamic tariffs and sharing of information

To further explore the smart metering data collected, additional exploratory analysis was carried out to try and establish trends. The first step of this analysis was to carry out crosstabulation of question 8 'SMART METER' against question 9 'THIRD PARTY'.

SMART METER * THIRD PARTY Crosstabulation

 

THIRD PARTY

Total

Strongly Agree

Agree

Unsure

Disagree

Strongly Disagree

SMART METER

Strongly Agree

2

5

6

19

7

39

Agree

4

10

15

21

2

52

Unsure

4

10

12

3

1

30

Disagree

4

2

0

1

0

7

Strongly Disagree

3

0

0

0

0

3

Total

17

27

33

44

10

131

Table 15

Table 15 shows the highest frequency of responses fall under the two categories highlighted. This suggests that users who 'Strongly Agree' and 'Agree' to having a smart meter installed, also 'Disagree' that recording information was a privacy concern. Although at this stage the result is not conclusive.

SMART METER * DYNAMIC TARIFF Crosstabulation

 

DYNAMIC TARIFF

Total

Strongly Agree

Agree

Unsure

Disagree

Strongly Disagree

SMART METER

Strongly Agree

11

12

7

8

1

39

Agree

8

22

12

10

0

52

Unsure

3

10

11

6

0

30

Disagree

1

0

1

4

1

7

Strongly Disagree

0

2

0

0

1

3

Total

23

46

31

28

3

131

Table 16

Table 16 compared SMART METER vs. DYNAMIC TARIFFS. The highest frequency of responses fell under the highlighted categories. This suggests that users who 'Strongly Agree' to 'Agree' about having a smart meter installed, also 'Agree' about the use of a dynamic tariff. With a trend beginning to emerge, it must now be established if the result is conclusive.

Table 17 - Pearson Smart Meter x Third PartyCorrelations

 

SMART METER

THIRD PARTY

SMART METER

Pearson Correlation

1

-0.495**

Sig. (2-tailed)

 

0

N

131

131

THIRD PARTY

Pearson Correlation

-0.495**

1

Sig. (2-tailed)

.000

 

N

131

132

**. Correlation is significant at the 0.01 level (2-tailed)

Table 18 - Pearson Smart Meter x Dynamic Tariff

Correlations

 

SMART METER

DYNAMIC TARIFF

SMART METER

Pearson Correlation

1

0.206*

Sig. (2-tailed)

 

.018

N

131

131

DYNAMIC TARIFF

Pearson Correlation

0.206*

1

Sig. (2-tailed)

.018

 

N

131

132

*. Correlation is significant at the 0.05 level (2-tailed)

Bivariate analysis was carried, in order to determine significance, magnitude and direction of the correlations discussed above. Firstly we shall look at the correlation of 'SMART METER' vs. 'THIRD PARTY'. The crosstabular analysis (Table 15) suggested that as users become more in favour of smart meters, they are less concerned about privacy. A correlation of -0.495 (Table 17), suggests that a moderate negative correlation exists between supporting smart meters and not having privacy concerns.

The second correlation tested was 'SMART METER' vs. 'DYNAMIC TARRIF'. From the crosstabular analysis (Table 16) it was established that a trend may exist, as users in favour of smart meters were also in favour of dynamic tariffs. Therefore we would assume a positive correlation between the two would exist. The correlation of 0.206 exists (Table 18), between users supporting smart meters and supporting dynamic tariffs, however this is weak.

To conclude, the analysis presented, suggests that it may be true that users who agree to have a smart meter installed will also support dynamic tariffs and sharing of information. However a larger sample size would confirm this.

Smart Home Design - Responses Graphs4.2.6 Smart Home Design Hypothesis

A) As energy users get older, they become less supportive of technology, such as smart metering and automation.

The one way ANOVA (Table 20) was used to test if a null hypothesis is significant. A significance of 0.351 is observed for 'SMART METER', as well as 'AUTOMATION' having a significance of 0.596. These both exceed 0.05 and therefore the null hypothesis is significant and should not be rejected. Our hypothesis of "As energy users get older, they become less supportive of technology, such as smart metering and automation." for our sample is therefore false and should be rejected.

Table 20 - Age Technology ANOVA

Table 19 - Age Technology Descriptive

4.2.7 Analysis of the open ended questions

Why do some energy users find it difficult to keep their energy bill to a minimum and others do not?

Common trends have been identified through analysis of responses to Question 7. Two categories emerged, the association was made between either a 'physical', such as the building, or 'human', the behaviour of the occupants. The responses are qualitative in nature and don't hold much statistical bearing, but however seek to gain insight into some of the issues surrounding home energy consumption directly from energy users. All of the responses can be found in Appendix M and have been used in the synthesis of the results. Total responses: 78

How does the trade-off between user control and automation make energy users feel?

Question 12 focused on automation, with an open ended question tagged onto the end. A common theme was established, that users were either in favour of, computer or human control. The analysis is qualitative in nature and does not hold a statistical representation. It is also limited due to the nature of the question, no particular technology is referred to therefore responses are broad. See Appendix M, for a copy of all the responses, as above the responses were used in the synthesis stage. Total Responses: 69

4.2.8 Additional comments from respondents

Question 16 allowed respondents to provide a comment, relating to any of the topics raised in the questionnaire. Any common themes have been identified and carried into the synthesis stage. The responses were broad in nature and have been analysed using keywords. A copy of all responses can be found in Appendix M. Total Responses: 20

4.2.9 Summary of open ended questions

In summary this section displayed some of the broader responses relating to home energy consumption, through three open ended questions. The limitations to this style of question raised in the methodology section, mostly related to the difficulty in analysis and interpretation (Fellows & Liu, 1997). However valuable insight has been gained which a series of rigid questions cannot (Fellows & Liu, 1997). The topics raised in the open ended questions, will now be carried forward into synthesis stage at the end of the results chapter. Results of the interviews carried out with homeowners will now be discussed.

4.3 Results of interviews and analysis

The results of the interviews has been presented in a table format, in order to summarise the transcribed interviews and provide ease of synthesis. A total of five participants took part in the interviews. Each of these transcribed interview can be found in Appendix J.

The first stage of the analysis was to create a summary table of some of the key concepts and answers which the interviewees discussed. The research instrument was created to be exploratory without too much emphasis on rigid responses. The analysis was go through the questions one at a time looking for trends, although at this stage of research the results are treated as indicative without much bearing due to the small sample.

4.3.1 Preliminary questions:

(A) Was any of the present information unclear, would you like to reread any sections?

Respondent A was the only interviewee which required the researcher to go through the information pack again. There was a particular struggle surrounding the terms "kWh", a somewhat foreign term to the respondent.

(B) How do you try and manage your energy consumption currently?

Two categories for managing energy consumption emerged during the interviews. It was either associated with behavioural activities such as "switching appliances off" or "turning off lights". Alternatively it was to do with the physical aspects of energy, such as "energy saving lights" or "Buying A rated appliances".

4.3.1 Device questions:

(A) Did you find the graphical or numerical feedback easier to understand?

The graphical feedback was praised for its ability to be able to absorbed at a glance, without need for further self calculation. As respondent A suggests "I can gauge how much (energy) I'm using and react to it.". Both respondent C and D understood the benefits of combining the graphical feedback with the numerical numbers of a smart monitor. As seen by respondent C's response, "at a glance you can see the graphical and the numerical gives you a bit more detail". This is built upon by respondent D's clarification of what he perceives the numbers to provide, "the most important number for people would be the cost.". Therefore the graphical feedback can be classed as a "trigger" for carrying out conservation actions. The "reward" comes in the form of the numerical feedback of reduced costs.

(B) Which of the five forms of feedback shown, do you feel would be most useful to reducing your energy consumption?

Appliance energy usage

Real Time consumption

Graphical

Cumulative Consumption

Numerical

Cumulative Consumption

Graphical

Real Time consumption

Numerical

Cumulative consumption numerical and real time consumption graphical were the two preferred forms of feedback. With three choosing real time consumption graphical and two choosing cumulative consumption numerical.

The real time consumption graphical was chosen by respondent A, B and C due to being able to visualise reductions. It provides the "trigger" effect where respondent C suggests she would "start switching things off" "if it's on red" meaning consumption was too high.

The cumulative consumption numerical was chosen by respondent D and E due to its relation with cost and cost reduction. Respondent D compared to miles per gallon in a car, which is often provided as an average. He suggests a "target in mind of how much you would like to spend that month", would be similar to trying to achieve low MPG. Both the respondents associated the cumulative numerical feedback as reducing cost, which respondent C also hinted to in her response, "show you how your bank balance is being spent.".

(C) Out of the three numerical forms of feedback, which one has the most value?

All five respondents chose the monetary value of energy usage. Due to this being expected, the questions was then asked why the other two were not so appealing. It is suggested by most of the responses that they were harder to relate to, as respondent A suggests they are "not actually telling me how much money is coming out of my pocket.". It is interesting that respondent D suggests "if you spending less money, then the other two are dealing with themselves, you'll be putting out less CO2 and using less kWh". Therefore primarily cost is what the respondents were most interested in, the CO2 and kWh were harder to relate too.

(D) Are there any features you would like to see on the energy monitor, that haven't currently been shown?

The most common response related to notification of energy waste from appliances on standby. Which both respondent A and Respondent D suggested. Respondent D was challenged for further insight as if the smart meter would already be able to tell you standby power at night with the real time consumption information. However he raised an important point that it should be able to remember previous days and provide the feedback at a later date. Another response related to patterns of energy consumption, respondent B suggests that he "would like it if it would highlight when there was any unusual energy consumption.

4.3.2 Energy management

(A) If you were informed that shifting your energy load from peak times (5pm -10pm) to non peak times would help you save money. Do you think a smart monitor would help you achieve this?

The monitor was mostly acknowledged that some of the features would help an energy user achieve this. Such as respondent E suggests, "you could make a note of when you are using it (energy) most and try and avoid those times.". Respondent D initially seemed sceptical that a smart monitor would help him achieve this, but realised he could compare different days consumption, to measure behavioural changes. This was also discussed by Respondent C who suggested behaviours would have to change, "but watching the TV you're not going to sit up until 2 in the morning just to save energy." suggests this would be selective. Respondent A interestingly suggested that her own common sense would be enough to not use energy at peak times.

4.4 Synthesis of the results

4.4.1 Energy Conservation

The categorisation of energy conservation, manifests itself as either 'Human' or 'Physical' (Wood & Newborough, 2003). The 'Human' element is the behaviour in daily life. Whereas 'Physical', Relates to the buildings we live in or appliances we own (Government, 2011). It is acknowledged by the government literature that reducing heat loss through fabric will ultimately have a greater impact than reducing appliance usage (DECC, 2012).

The Human Aspect

The human aspect of energy consumption relates to the occupants and how they interact with their surroundings. The majority of the sample of this thesis felt they were concerned about energy conservation within their home. The works of Brandon & Lewis (1999) and Hargreaves et al (2010), highlighted how money was the primary motivation towards energy conservation. The results of Question 6 concluded that 74% of respondents chose money as the primary motivation, followed by 20% claiming environmental impact. The number of respondents associating environmental impact with motivation for reduced energy consumption is somewhat surprising. As Wilhite & Ling (1995) claim the link between environmental impact and energy consumption is weak, suggesting attitudes may be changing.

It was discovered that the student group, were the least concerned about energy conservation. The Powering the nation report (2012), highlighted how multi-households with no dependents used the most energy in terms of washing and drying. Further research into households consisting of multiple individuals, may gain insight into why higher energy consumption and less concern for conservation occurs. Multiple occupancy home environments, often have difficulties in reducing their energy consumption, due to conflicting views of occupants. As decisions of individuals are often influenced by that of others (Cialdini, 2007), differences in occupancy behaviour (Stevenson, 2010), (Gentoo, 2010), (Gill et al., 2010), will impact the amount of energy being used as a whole.

The acknowledgement that leaving lots of appliances on standby was made some of the respondents of the questionnaire. The literature suggests that cheap energy prices of the past have lead to the development of wasteful habits (Wilhite & Ling, 1992). With ownership of appliances being at some 42 on average, contributing to 9-16% of a home's energy bill, there must be scope for improvements (Energy Saving Trust, 2012).

Understanding appliance energy usage was also raised as a difficulty. Participants in the Home, Habits and Energy study (2010) also had very little understanding of the cost of each appliance. The creation of awareness of how different types of technology use energy will be important for reducing consumption (Gyberg & Palm, 2009).

The final human aspect to be discussed relates to the requirement of living in a warm environment. Some responses related to the need for a warm lifestyle. This may be either required due to poor health or due to a comfort requirements. It is acknowledged in the study Reducing Energy Consumption (1999), that comfort takes precedence over cost. Cultural aspects also play into whether a warm lifestyle is preferred or not, as was found in the study by Wilhite & Ling (1992). Therefore benefits achieved through improving home efficiency, are likely to be either comfort or cost related (Clinch & Healy, 2001).

The Physical Aspect

Limitations to effective efficiency improvements, through modifications in behaviour, often manifest themselves through the physical home environment (Shove, 2003). The Energy Efficiency Strategy (2012), outlined that 45% of properties in the UK were constructed before World War 2 (DECC, 2012). Therefore there must be considerable scope for improvements to be focused on in this area, before users are forced into making major behaviour modifications.

Respondents, associated the building they live in as either an advantage to being energy efficient or a disadvantage. For example respondent 92 made the comment, "the house is quite old and I know it is not designed, built or insulated with energy saving in mind.". This therefore constitutes a barrier to the uptake of 'human' energy saving practice before a 'physical' solution is found. However some respondents, such as respondent 40, having already installed double glazing and loft insulation, suggesting doing anything else within their lifetime would not provide financial returns.

Another aspect was raised by respondent 79, who claimed living in poorly insulated rented accommodation was contributing to their energy bill. However due to not being the homeowner felt unable to remedy the situation. This is also discussed in Energy conservation programs for consumers, suggesting occupants of rented accommodation do not have the ability to implement energy saving technologies (Bernward & Muller, 1983).

An association was discovered, that limitations between heating required and efficiency of its operation exist. Limitations in the plant currently owned or in some cases the lack of gas availability forced users to rely on electricity. The lack of efficiency of central heating plant owned by energy users should clearly supersede the requirement to modify behaviour. According to The carbon plan for every 1°C reduction in heating, 10% saving can be achieved (Government, 2011). Therefore such schemes aiming to provide future investment in home improvements, must be built upon and delivered before behavioural changes take place.

When exploring appliance efficiency, 84% of respondents 'Agreed' to 'Strongly Agreed' that energy efficiency rating of appliances was important to them. The incentive to buy a cheaper model appliance over a more energy efficient one, is due to increased cost (Wood & Newborough, 2003), (Money supermarket, 2013). Limitations of the context of this question therefore provided misleading results, as the association between efficiency and increased capital cost was not made. As respondent 6 suggests, the capital cost of purchasing new appliances has to be justified, with an example of "C = £199 and B = £299". Therefore without the justification on the purchase of a more efficient appliance, energy users are unlikely to do so.

4.4.4 The Introduction and Operation of Smart Metering

Smart metering was first discussed under the carbon plan (2011), setting out a mandatory rollout of smart meters by 2019, as well as the provision of in home smart monitors (Government, 2011). With energy demand set to increase 30-100% by 2050 (DECC, 2012), focus must be put on a flexible, smart and responsive energy system (DECC, 2011). This seeks to be achieved through DSM, with smart metering being the enabler (DECC, 2011). The confusion within the domestic sector as to what constitutes a smart meter or a smart monitor, can be influential when trying to collect data on opinions (Which?, 2013), (DECC, 2013).

Considering those answering 'Strongly Agree' and 'Agree' to question 8, as being supporters of smart meters, and those answering 'Unsure', 'Disagree' and 'Strongly Disagree' as being non supporters. Of the respondents, 70% would be supporters and only 30% would be non supporters. Confusion may have arisen, as to what constitutes a smart meter (Which?, 2013), (DECC, 2013), however description was provided in the question. Comparing my results with the DECC quantitative study (2013), my results favour the installation of a smart meter, whereas only 40% of their sample favoured (DECC, 2013). This deviation can most likely be associated to the lack of sample diversity, as it is said that a strong correlation exists that younger larger households being more supportive (DECC, 2013).

Generally, users within the sample were unsure as to whether third parties recording consumption information was a privacy concern. Through analysis of the data, a trend was discovered that users who fell into the supporter category of smart metering, also were to some extent in favour of dynamic tariffs and sharing of information. The collection and dissemination of information is required in order to be able to operate future smart grids (Rajagoplana et al., 2011). The requirement will be put onto consumers, to work with utility companies in order to manage energy consumption (McDaniel & McLaughlin, 2009), therefore having inconclusive results on users neither supporting nor being against sharing of information requires additional research.

Much negativity around increasing energy bills and vested interests by energy suppliers wass raised. This may account for why respondents felt unsure as to whether a dynamic tariff would help them save money. As the study "Unlocking the €53 billion savings from smart meters in the EU" suggests there is a current lack of supplier offer and customer acceptance. Respondent 29 felt that the complexity of energy prices, promoted a desire for one tariff for each supplier. However rate adjustment to meet demand was supported by respondent 102. Consumer opinions surrounding DSR through the use of tariffs is generally seen as positive (DECC, 2012). The results of this thesis however, indicate it is inconclusive if energy users feel it would help them save money or not.

4.4.5 Smart Monitors and Smart Home Design

The preferred form of feedback for home energy usage, was firstly personal computer/laptop based feedback, at 62% of responses. Following this was smart phone based feedback, with 22% of respondents selecting this option. It is said that for feedback to be effective it must be provided directly after an action (Stern & Arsonson, 1984).

However feedback provided via a computer is unlikely to occur directly after an action takes place. It is also surprising that only 10% of the sample chose a dedicated device. Being as the primary form of home energy feedback being supplied by energy companies (British Gas, 2009), (EDF Energy, 2013), (Eon, 2012), (Npower, 2010), (Southern-Electric, 2013), (Green energy options, 2010).

Providing disaggregated feedback on consumption, will allow users to take control of their own energy usage (Gyberg & Palm, 2009). As was highlighted in the study "Effects of Feedback on Residential Electricity Consumption", cost based feedback with breakdowns of consumption, aided in reducing energy consumption (Farhar & Fitzpatrick, 1989). It is therefore beneficial to the design of future feedback devices, that the majority (84%) of the sample felt this information would be useful.

Comparing consumption information with neighbours, has shown to have influence on changing user behaviour (Stern & Arsonson, 1984), (Pallak et al., 1980), (Arvola et al., 1993), (Ueno et al., 2006). The acceptance of this however, is not conclusive within the sample of this dissertation. Analysis of question 15 of the questionnaire, would indicate clarification need to be delivered towards the execution of sharing of information, as the central tendency was unsure.

The provision of disaggregated feedback on appliances was generally accepted by the sample. Being specific to an energy users environment was highlighted by "Reducing Household Energy Consumption", where feedback which was not personal to themselves was seen as a drawback (Brandon & Lewis, 1999). The "Home, Habits, and Energy" study highlighted how providing cost information about consumption, rarely altered appliance usage patterns afterwards (Pierce et al., 2010). There is still a challenge, as to how presenting this information should be considered.

4.4.6 The effectiveness of smart energy monitors and future design considerations

The literature review of smart monitors, was carried out on six of the major energy companies. It is suggested by "which?", that companies supplying smart monitors are pairing them with higher energy tariffs (Which?, 2013). It is the governments vision to provide all homes with a smart meter and smart monitor, under the carbon plan (Government, 2011). However this cost should not be at the expense of the consumer.

The final question of the questionnaire, allowed respondents to freely raise issues about the topics discussed. A common response was negativity around energy suppliers and the "vested interest" for commercial gain. In order for energy monitoring and the installation of smart meters to be sustainable, the benefits of such schemes has to feed through to consumers. Purely relying on increased market competition as a means of controlling cost, is highlighted by "which?" as a dangerous strategy, where cost will most likely fall on the consumer (Which?, 2013).

The identification of the main features of current smart monitors was made within the literature review. Interviews were then carried out to gain additional insight. Firstly it was important to discover what the interviewees currently did to try and manage their energy consumption. As expected the two categories of either 'human' or 'physical' emerged. It was determined through analysis of the smart monitor manuals and the energy literature, that feedback is provided in graphical or numerical format.

The graphical feedback was discussed by respondents, as a means to 'trigger' a reaction. The literature suggests for feedback to be effective, it must be provided after an action has taken place (Stern & Arsonson, 1984). The "Making energy visible" study highlighted how the use of visual cues, such as a high meter reading, would spark a reaction (Hargreaves et al., 2010). Even at the simplistic level, such as in "Effective dissemination of energy-related information", where signage by lighting reminded people to turn off lights (Dennis et al., 1990), shows how feedback following an action is effective (Stern & Arsonson, 1984).

The review of current smart monitors identified two types of graphical feedback. The real time consumption graphical was the most preferred form of feedback in the interviews. Mostly due to the ability to watch the bar fluctuate in real time, providing instant feedback for any conservation techniques being carried out. The results of "Dynamic energy-consumption indicators for domestic appliances", claim 10-20% reduction in energy consumption, through a use of the combination of graphical and numerical feedback (Wood & Newborough, 2003).

However as highlighted in the smart monitors section of this thesis, current real time consumption increments, are being set too high (as high as 22kW), therefore poses a major drawback to providing accurate feedback (British Gas, 2009), (Npower, 2010), (Green energy options, 2010). Therefore to rectify this, a profile must be established of the home which a monitor is installed, basing the speedometer readings on an average. A feature that was lacking from current smart monitors.

The numerical feedback was described by interviewees as a means of 'reinforcing' what the graphical feedback was telling them. Presented mostly through cost information, described as the reward for carrying out energy saving actions. The results of the questionnaire, and much of the energy literature, also associates cost as the primary motivator for reducing energy consumption (Brandon & Lewis, 1999), (Hargreaves et al., 2010). The Cumulative consumption numerical, was chosen by interviewees, as it could be used as an aid for budgeting. Many of the current smart monitors require you to enter your own tariff information. This may however, contribute to complexity of operation, as users in the "Making energy visible" study, often failed to enter this information (Hargreaves et al., 2010).

Additionally to cost information, smart monitors provide numerical feedback in the form of kWh and CO2. This information is generally seen as hard to relate too (Hargreaves et al., 2010), with respondent D of the interviews claiming they will deal with themselves. This is somewhat true, however cost of fuel does not move in unison with the amount of kWh consumed or CO2 produced. Arguable it is the kWh that should be the primary focus, as this can be measured within a home, whereas cost and CO2 are an external factors

Focused on the trade off between increased energy efficiency (through automation) and reduced user control, generally was seen as 'Positive'. Time constraints, human error and increased efficiency were some of the topics raised by respondents. However the negative aspects, such as, frustration with modern technology, disconnection with the occupants (who may have health issues) and increased complexity, were some of the issues raised.

5.0 Conclusion

5.1 Introduction

This overall aim of this research was to advance understanding in the field of Smart Home Technologies. Predominately through research into user behaviour, smart metering and smart monitoring. This chapter aims to conclude all of the findings of the research, making recommendations towards future research.

5.2 Research Objectives: Summary of Findings and Conclusions

The focus of this research centred around Four research objectives, the final of which shall be achieved in the recommendations section. A summary has been produced of key findings for each research objective.

1) Establish the barriers to reduced energy consumption

The 'Human' element of energy consumption, creates several challenges which need to be overcome. It was determine that decisions are either 'Conscious' or 'Unconscious'. Decisions made by energy users mostly fall under 'Unconscious'. The lack of visibility of energy consumption, makes it challenging to recognise where energy is being consumed. Receiving bills as an accrued total each month, hinders ability to gain understanding about consumption. Therefore highlighting a requirement for disaggregated feedback. Difficulties in understanding energy consumption, can also manifest themselves through the socio-demographics of a household. The cost vs. comfort trade-off which varies at different levels of income influences the ability provide feedback on reducing consumption. Finally the influence of the fallback effect, will reduce effectiveness of implemented smart monitors.

2) Explore wider user opinions related to issues of energy consumption and smart home technologies

It can be concluded that, energy users are generally concerned about energy consumption within their homes. Barriers to reducing consumption, manifest themselves through both 'human' and 'physical' avenues. The motivation for reducing energy consumption, presents itself through 'Cost', followed by 'Environmental Impact'. Generally there is support for smart metering, however it is inconclusive whether sharing of information or third part tariffs is supported. Feedback on appliances, will be a useful tool for managing home energy usage, but the medium to present this information is still unclear. Finally the issue of multi-occupancy households and energy management needs to be tackled.

3) Identify key technologies being supplied by energy suppliers and review their potential for saving energy

The review carried out of 6 smart monitors provided by energy companies, highlighted how there is significant shortfalls of current design. The feedback provided, must facilitate reduction of wasteful behaviour, stemming from the learning stage (Appendix F). The works of Hargreaves et al (2013) suggests preventing to do so, will have the revese effect of reinforcing current wasteful behaviour. The current focus on the financial benefits of smart monitors, hinders the effectiveness of the feedback. Emphasis on reducing consumption through learning about kWh patterns, then providing the rewards as cost feedback at a later date, may be beneficial. As the current rise in costs hinders the learning process, whereas the cosntant variable of kW consumption may be easier to recognise.

5.3 Recommendations

The final research objective, was to formulate recommendations for future smart monitor design, relating to energy conservation. Throughout carrying out this research, it was apparent there is a significant lack of literature surrounding smart monitors. The first recommendation, is to suggest a thorough review is carried out to determine the long term impact of current smart monitors. The impact of the fallback effect, seriously hinders the ability for devices to be effective.

There is a current lack of understanding by energy users, relating the consumption of power (kW) with reduced energy usage. Emphasis on altering behaviour, is focused on through feedback in the form of cost. This is unreliable, as the fluctuating price of energy can mask the conservation actions carried out. If the association is made between level of power consumed and reducing energy costs, it is likely to create more sustainable changes. The introduction of DSR through dynamic tariffs, will therefore yield greater benefits to users who understand their energy load profile.

Finally, clarity needs to be delivered to consumers as to the different purposes smart meters and smart monitors serve. Currently banding them under the same category confuses consumers, with energy companies possible gaining a commercial advantage. The benefits of DSM must feed through to consumers, through the implementation of DSR as the reward.