Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Do Job Groups Influence Email Usage in A Similar Way As
Roles Do? A Study of Academics
Benjamin M. Silverstone
The influence of roles in the use of email has been explored at the meso-cultural level where
senior management, middle management, business support and academic roles were identified
and profiles created. It was observed that significant differences existed between the roles in 9
out of 12 markers. In the same study, conducted by Silverstone (2014) it was suggested that the
same patterns may exist at the micro-cultural level.
This paper seeks to explore the perception that the same patterns may exist at the micro-cultural
job level. Using data gathered by Silverstone (2014) the Academic role has been broken down
into 4 appropriate job groups and an unspecified academic group.
From the original 1010 responses gathered by Silverstone (2014), 481 fell into the Academic role
and were used for this study. Differences between the groups were explored using descriptive
and statistical methods and some differences were observed between the different roles being
explored. Analysis was conducted descriptively and analytically with Chi-square, ANOVA and
bivariate Correlation being used for statistical analysis of quantitative components and extensive
coding and content analysis used for qualitative components.
The findings demonstrated that there were no significant differences between email usage at the
job level when compared to the differences meso-cultural role level. A total of 12 markers were
tested statistically. Differences in the correlation Maximum manageable for sent and received
loads were significant as were differences in perceptions of wastage. At the role level, 9 of the 12
markers showed significant differences, at the job level only 2. The findings are significant as they
help to validate that the predominant indicator of differences in email use is role as opposed to
organisational culture.
JEL Codes: M15, M20, M54
1. Introduction
Work by Silverstone (2012) suggested that there may be a strong link between effective use
of email and the relationship between the users involved in the communication process. The
theory used cultural markers as a basis upon which to suggest that a relationship based
upon mutual cultural understanding would enhance the use of email. This relationship
construct was termed subjective distance and suggested that a less rich means of
communication would become richer where a strong relationship is present between those
communicating.
The influence of culture was further explored by Silverstone (2014) who focused on the
influence of role culture in the email process. Identified roles were defined within the sector
and this information was used as a dependent factor in the testing of wide variety of email
use variables. It was found that significant differences did exist between the roles related to
the way that they use email. This helped to suggest that the earlier assertions about cultural
impacts on mail usage carried weight.
Whilst evidence is present to support the idea that role culture does have an Impact on
email usage, Silverstone (2012) suggested that these similarities would exist more strongly
________________________________________________________________________
Mr. Benjamin M. Silverstone, School of Applied Computing, Faculty of Architecture, Computing and
Engineering, University of Wales Trinity St David, Swansea. Email: Benjamin.silverstone1@sm.uwtsd.ac.uk
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
at the micro-cultural level between individuals functioning in the same department or
working group. These assertions were based upon the findings of Huczynski and Buchanan
(2001) and Trice and Beyer (1984) who suggested that shared beliefs and values as well as
history, ceremonies and rites were strong influencers on communication.
The premise upon which Silverstone (2012) built this assumption was that these factors
would be most strongly realised in smaller groupings where the driving factors of cultural
development would be most strongly represented. This paper seeks to explore this assertion
in micro groupings of Academic staff across the Welsh FE sector.
2. Method
Use was made of the primary research conducted by Silverstone (2014). The same data set
was used and the responses from the Academic Role were filtered to produce the data set.
A total of 480 response records were isolated from the main data set. Within this, 40
different job groups were identified. A number of these were represented by too few
respondents to enable meaningful analysis.
In order to strengthen the analysis it was decided to combine the job groups along the broad
academic groupings used within the sector. This significantly increased and balanced the
response numbers within the groups to enable meaningful analysis. Figure 1 below
illustrates the groups and representation within the data set. The largest group included was
the unspecified academic group. In order to mitigate for this imbalance, statistical analyses
were carried out with and without this group included.
The structure of the analysis mirrors that used by Silverstone (2014) in order to provide a
direct comparison.
Figure 1 – Breakdown of data set by specialisation groupings.
Group
Number of Respondents
Proportion of Data Set
Health, Care, Services 63
13.1%
and Independent Living
Skills
Humanities
87
18.1%
Information
Technology 71
14.8%
and Essential Skills
Science, Engineering and 71
14.8%
Construction
Unspecified Academic
188
39.2%
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
3. Results
Drawbacks and benefits related to the use of email were asked for through an open ended
question. In each case 480 responses were received. These were coded and the following
themes emerged as most popular. In terms of drawbacks, Lack of human interaction, the
potential for misinterpretation and the possibility of damaging messages were the key
themes identified. The benefits identified were speed, reliability and ease as well as the
record of messages for some of the employment groups.
Sent and received message load along with the perception of change are shown in figures 1
to 4 below. For analytical purposes the number of categories was collapsed to enable a
valid chi square test to take place.
Figure 2, the perceived changes in received message load by Academic Employment Group
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
2
8
10
10
14
44
Count
Stayed the% within New
same
employment 3.2%
9.3%
14.3%
14.1%
7.5%
groups
[10] How has
Count
58
77
57
59
165
the volume of
received
Increased % within New
messages
employment
93.5%
81.4% Employment
83.1%
88.7%
Figure
1, the perceived changes
in sent
message 89.5%
load by Academic
Group
changed
in
groups
recent years?
Count
2
1
3
2
7
%
within
New
Decreased
Health,
Humanitie Information Science,
Unspecifie
employment 3.2%
1.2%
4.3%
2.8%
3.8%
Care,
s
Technolog
Engineering
d
groups
Services
y 70 and and
Academic
Count
62
86
71
186
and
Essential
Constructio
%
within
New
Total
Independen
employment
100.0%
100.0% Skills
100.0% n 100.0%
100.0%
t
Living
groups
Skills
Count
57
%
within
Increased New
90.5%
employmen
[8] How
t groups
has the
Count
2
volume of
sent
within
Decrease %
message
New
d
3.2%
s
employmen
changed
t groups
in recent
Count
4
years?
%
within
Stayed
New
the same
6.3%
employmen
t groups
Count
63
%
within
Total
New
100.0%
employmen
t groups
9.3%
416
87.6%
15
Total
3.2%
475
100.0%
71
56
53
157
394
83.5%
80.0%
74.6%
86.3%
83.7%
1
2
2
5
12
1.2%
2.9%
2.8%
2.7%
2.5%
13
12
16
20
65
15.3%
17.1%
22.5%
11.0%
13.8%
85
70
71
182
471
100.0%
100.0%
100.0%
100.0%
100.0
%
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Figure 3, number of messages sent per day by Academic Employment Group
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
38
40
33
49
94
254
Count
0 -% within New
10 employment 60.3%
groups
Count
16
11
%
within
New
20 employment 25.4%
groups
Count
5
21
- % within New
30 employment 7.9%
groups
[7]
On Count
4
average, how31
many emails- % within New
do you send40 employment 6.3%
groups
in a day?
Count
0
41
%
within
New
50 employment 0.0%
groups
Count
0
71
- % within New
80 employment 0.0%
groups
Count
0
81 % within New
+ employment 0.0%
groups
Count
63
% within New
Total
employment 100.0%
groups
46.0%
46.5%
69.0%
50.0%
52.9%
31
25
16
69
157
35.6%
35.2%
22.5%
36.7%
32.7%
11
6
6
18
46
12.6%
8.5%
8.5%
9.6%
9.6%
4
4
0
4
16
4.6%
5.6%
0.0%
2.1%
3.3%
1
3
0
1
5
1.1%
4.2%
0.0%
0.5%
1.0%
0
0
0
1
1
0.0%
0.0%
0.0%
0.5%
0.2%
0
0
0
1
1
0.0%
0.0%
0.0%
0.5%
0.2%
87
71
71
188
480
100.0%
100.0%
100.0%
100.0%
100.0%
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Figure 4, Number of messages received per day by Academic Employment Group
Count
0 -% within
10 employment
groups
Count
11 -% within
20 employment
groups
Count
21 -% within
30 employment
groups
Count
31 -% within
[9] On average,40 employment
groups
how many emails
do you receive in
Count
a day?
41 -% within
50 employment
groups
Count
51 -% within
60 employment
groups
Count
61 -% within
70 employment
groups
Count
81 + % within
employment
groups
Count
% within
Total
employment
groups
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
22
15
17
20
36
110
New
35.5%
17.4%
23.9%
28.2%
19.3%
23.1%
19
36
27
25
84
191
30.6%
41.9%
38.0%
35.2%
44.9%
40.0%
13
18
12
17
34
94
21.0%
20.9%
16.9%
23.9%
18.2%
19.7%
5
8
8
7
15
43
8.1%
9.3%
11.3%
9.9%
8.0%
9.0%
2
7
3
1
10
23
3.2%
8.1%
4.2%
1.4%
5.3%
4.8%
1
2
1
1
5
10
1.6%
2.3%
1.4%
1.4%
2.7%
2.1%
0
0
0
0
1
1
0.0%
0.0%
0.0%
0.0%
0.5%
0.2%
0
0
3
0
2
5
0.0%
0.0%
4.2%
0.0%
1.1%
1.0%
62
86
71
71
187
477
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
New
New
New
New
New
New
New
New
The average time spent on email usage for each of the employment groups was gathered
using a sliding scale from 0 to 180 minutes. HCSILS spent on average 50.73 minutes per
day, Humanities spent 59.23 minutes per day, ITES spent 44.66 minutes per day, SEC
spent 38.68 minutes per day and the unspecified academic group spent 49.42 minutes per
day.
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
The number of messages that users perceived were manageable to send and receive in a
day was gathered using an open ended text box. HCSILS perceived that an average of
15.52 messages could be sent and 14.80 could be received. Humanities perceived that an
average of 14.34 messages could be sent and 13.98 could be received. ITES perceived that
an average of 11.89 messages could be sent and 12.40 could be received. SEC perceived
that an average of 8.82 could be sent and 24.80 could be received. UA perceived an
average of 17.29 could be sent and 11.71 could be received.
Figure 5 below illustrates the proportions of each group that wished to change their email
usage. This was followed up by an open ended question asking respondents to justify their
answers. For those who did wish to change their usage the key reasons reported were
volume and content management and a desire to receive fewer unsolicited emails. For
those who did not wish to change their usage, respondents generally believed that the
current levels of usage are manageable but should not increase.
Figure 5, the desire to change email usage by Academic Employment Group
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
21
38
23
24
78
184
Count
%
within
YesNew
33.9%
employment
[13]
Would
groups
you like to
change your
Count
41
email usage?
%
within
No New
66.1%
employment
groups
Count
62
%
within
Total
New
100.0%
employment
groups
44.2%
32.4%
33.8%
41.9%
38.7%
48
48
47
108
292
55.8%
67.6%
66.2%
58.1%
61.3%
86
71
71
186
476
100.0%
100.0%
100.0%
100.0%
100.0%
Respondents were asked whether they consider others before sending emails. The results
are shown in figure 6 below .Respondents were also asked to justify their responses. The
main issues reported were that users tended to consider their own and others’ time
management and expectations as well as the appearances and interpretation of the
messages they were sending.
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Figure 6, consideration of others when sending emails by Academic Employment Group
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
46
78
59
56
160
399
Count
%
within
[14]
In
general,
doYesNew
73.0%
employment
you consider
groups
the impact on
the recipient
Count
17
before
%
within
sending
No New
27.0%
emails?
employment
groups
Count
63
%
within
Total
New
100.0%
employment
groups
89.7%
83.1%
78.9%
85.1%
83.1%
9
12
15
28
81
10.3%
16.9%
21.1%
14.9%
16.9%
87
71
71
188
480
100.0%
100.0%
100.0%
100.0%
100.0%
Respondents were asked to report whether they waste any time when using email. The
results of this can be seen in figure 7 below. Those who answered yes to this were then
asked to provide an estimate of what proportion of time they spend using email is wasted.
HCSILS reported an average of 17.46% wastage. Humanities reported an average of
14.76% wastage. ITES reported an average of 14.54% wastage. SEC reported an average
of 16.03% wastage and unspecified academics reported an average of 14.13% wastage.
The main reasons for wasted time provided by respondents was the issue of work related
emails that are not relevant or have been duplicated. Respondents were also asked to
identify behaviours from a list established through review of existing literature (figure 8
below).
As well as identifying behaviours, users were asked to select the one that they thought was
most important and provide an example. There was a fairly even spread across the issues
with content that is not relevant being the most common. Poorly written, aggressive in tone
and the same message, containing the same content from multiple sources were also
reported on.
Figure 7, the perceptions of wasted time when using email by Academic Employment Group
Total
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
24
21
39
24
68
176
Count
%
within
No New
38.1%
employment
[17] Do you
groups
waste
any
time
using
Count
39
email?
%
within
YesNew
61.9%
employment
groups
Count
63
%
within
Total
New
100.0%
employment
groups
24.7%
55.7%
33.8%
36.4%
37.0%
64
31
47
119
300
75.3%
44.3%
66.2%
63.6%
63.0%
85
70
71
187
476
100.0%
100.0%
100.0%
100.0%
100.0%
Figure 8, wasteful behaviours by Academic Employment Group
Response
Health,
IT,
Care,
Humanities Essential
Service,
Skills
ILS
Inappropriate content
12.7%
13.8%
8.5%
Aggressive tone
30.2%
28.7%
25.4%
Bullying
20.6%
9.2%
7.0%
Content you found offensive
7.9%
13.8%
7.0%
Sent by the sender to avoid F2F 52.4%
56.3%
45.1%
contact
Poorly written
66.7%
70.1%
66.2%
Hastily composed without due 57.1%
64.4%
54.9%
consideration
Content that is not relevant to you
81.0%
89.7%
77.5%
Same message from multiple 65.1%
77.0%
60.6%
sources
None of the above
6.3%
4.6%
11.3%
Science,
Unspecified
Engineering,
Academic
Construction
9.9%
33.8%
16.9%
8.5%
53.5%
17.6%
28.7%
13.3%
12.2%
53.7%
74.6%
57.7%
66.0%
56.4%
87.3%
64.8%
82.4%
67.6%
2.8%
6.9%
Figures 9 to 11 below illustrate the responses to questions about attendance at training. The
questions focused on whether users had attended training in the past 12 months, the nature
of the training, whether it was relevant and if not, why not. The information on why training
was not relevant was gathered using an open ended question and the main themes were
that the training was not relevant to the employment group, it was not required or it failed to
achieve the stated aims.
Figure 9, the attendance at email training within the past 12 months by Academic Employment
Group
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
12
9
8
8
28
65
Count
%
within
[22] Have youYesNew
19.0%
employment
attended
groups
training on the
use of email
Count
51
in the past 12
%
within
months?
No New
81.0%
employment
groups
Count
63
%
within
Total
New
100.0%
employment
groups
10.3%
11.3%
11.3%
15.1%
13.6%
78
63
63
158
413
89.7%
88.7%
88.7%
84.9%
86.4%
87
71
71
186
478
100.0%
100.0%
100.0%
100.0%
100.0%
Figure 10, whether the training attended was relevant to the respondent by Academic Employment
Group
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
Count
12
9
7
7
28
63
%
within
YesNew
63.2%
42.9%
46.7%
35.0%
52.8%
49.2%
employment
[22b] If you
groups
did
attend
training, was it
Count
7
12
8
13
25
65
appropriate?
%
within
No New
36.8%
57.1%
53.3%
65.0%
47.2%
50.8%
employment
groups
Count
19
21
15
20
53
128
%
within
Total
New
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
employment
groups
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Figure 11, the nature of the training undertaken by Academic Employment Group
Count
Software
or% within
hardware training employment
groups
Count
Content
% within
[22] Have youmanagement
employment
attended trainingtraining
groups
on the use of
email in the past
Count
12
months?Accredited
% within
[other]
course
employment
groups
Count
% within
Other training
employment
groups
Count
% within
Total
employment
groups
Total
Health,
Humanities Information Science,
Unspecified
Care,
Technology Engineering Academic
Services
and
and
and
Essential Construction
Independent
Skills
Living Skills
7
7
5
5
12
36
New
70.0%
87.5%
83.3%
100.0%
66.7%
76.6%
1
0
0
0
2
3
10.0%
0.0%
0.0%
0.0%
11.1%
6.4%
1
0
0
0
2
3
10.0%
0.0%
0.0%
0.0%
11.1%
6.4%
1
1
1
0
2
5
10.0%
12.5%
16.7%
0.0%
11.1%
10.6%
10
8
6
5
18
47
100.0%
100.0%
100.0%
100.0%
100.0%
100.0%
New
New
New
New
4. Discussion
These results provide us with an insight into email user profiles of Academic employment
groups at the micro-cultural level. This discussion will focus upon identified groupings within
the Academic role identified by Silverstone (2014).
When considering the drawbacks and benefits of email usage it can be seen that, whilst the
percentages are not high, the HCSILS and Humanities groups are most likely to perceive a
lack of human interaction as a drawback. Silverstone (2014) identified this as an observation
made by the senior management and business support roles. The use of face to face
communication has been identified as a method of improving communication effectiveness
within the organisation by improving relationships (Silverstone 2012).
There may be a link with the nature of the subjects being taught by academics within these
groups. For example, the HCSILS group would include subjects such as Effective
Communication, Psychology and Personal and Professional Development which would
heavily stress the importance of good communication. The humanities group would include
subjects such as People Management, Tourism Management, English and Sociology which
again would consider communication issues more deeply. A number of courses taught
within Humanities are used for the up skilling of the Business Support role as well as
preparation for management roles.
Excessive load was also identified by a significant minority of respondents in all of the
groups with the exception of HCSILS. Silverstone (2014) linked this consideration to usage
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
and perceived manageability. However, in this case, despite HCSILS being least likely to
identify excessive load as a drawback, this group is most likely to identify that both sent and
received message load have increased. Additionally, they do not have the lowest perception
of manageability either. This suggests that observations at the meso-cultural level do not
hold true at the micro-cultural level.
The final most commonly reported drawback is that of the potential for damaging messages
and misunderstandings. This was not identified as a significant concern at the meso-cultural
level but the basis for it is just. Authors such as Miller (1999), Cunningham and Greene
(2002) and Kierkegaard (2005) discuss the potential issues associated with damaging
messages which include messages that can be sent to a large number of people and then
forwarded on again which may damage reputations as well as the possibility of the actions
of an individual causing damage to the organisation. Inclusion of this as a concern shows a
consideration of others and the potential impact of email usage.
In terms of the key benefits, all groups identify that speed, reliability and ease are the key
benefits of email communication. There may be a link with the nature of academic groups
where there is less time spent in an office with direct access to email. As a result, an email
can be sent when it is convenient and then actioned when appropriate. This fits with
identification of the asynchronous nature of email with the associated personal time
management benefits also being present. This benefit was strongly identified by all roles
studied by Silverstone (2014) which suggests that regardless of the cultural level, the
perception of speed, reliability and ease is strong.
Speed and cost benefits have been key factors in the rapid adoption of email (Sillince et al
1998) and this view appears to permeate regardless of other issues such as a perception of
increase in usage or disparity between actual and perceived manageable usage statistics.
Significant minorities of respondents from across the groups, especially HCSILS and SEC,
identified the record of messages as a benefit. This benefit was observed by Silverstone
(2014) within the overall Academic role and it now appears that HCSILS and SEC group are
the main cause for this. Storage of messages can be used to protect individuals in cases of
bullying or other inappropriate usage (Collins 1986 and Seshadri and Cartenson 2007).
However, StAmant (2001) showed that despite being written and recorded, this did little to
deter bullying via email.
The benefit of a written record may also be linked to the ability to keep a written record of
conversations or meetings. As academics spend the majority of their time teaching, the
facility to refer back to something at a later date may be a benefit. The ability to refer back is
also cited by Collins (1986) and Seshadri and Cartenson (2007). Despite the potential
benefits, there is no evidence to suggest why HCSILS and SEC groups most strongly
identify this benefit.
When considering sent and received messages, whilst differences can be observed in the
descriptive statistics, the differences were not considered to be statistically significant
according to Chi-square testing, therefore there is no relationship between micro-cultural
academic group and sent (x=15.235, p=0.229) or received messages load (x=8.912,
p=0.179).
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
The same analysis was carried out with the unspecified academic group excluded in order
to assess the differences in the identified academic employment groups. Chi-square testing
showed that there is still no statistical difference between the groups in terms of both sent
(x=8.912, p=0.179) and received message load (x=14.014, p=0.122). Whilst the influence of
employment group is still not statistically significant, there appears to be greater significance
with the unspecified academic group removed.
Descriptive analysis shows very little difference in sent message load. The majority of
respondents for all groups sent in the 0 – 10 messages per day category with a sharp
decline thereafter. This pattern follows the overall observations for the Academic role
identified by Silverstone (2014). These observations would account for the lack of statistical
difference in group behaviour and also lend weight to the argument that statistical difference
does not extend beyond the meso-cultural level.
The average sent message load for each group is as follows, HCSILS sent an average of
11.83 messages daily. Humanities sent an average of 13.66 messages daily. ITES sent an
average of 14.392 messages daily. SEC sent an average of 9.79 messages daily.
Unspecific Academics sent an average of 12.96 messages daily. There are slight
differences with the SEC group sending the lowest average number of messages daily and
the ITES group sending the most. Silverstone (2014) identified a mean sent of 12.70
messages daily for the Academic Role.
Further descriptive analysis of received message load shows a similar pattern as observed
in the Academic role by Silverstone (2014). There is one exception. The HCSILS group has
a slightly different pattern, receiving a much higher proportion of messages in the 0 – 10
category than the other groups. The produces a pattern of reducing numbers in each
category of received messages where the other groups peak in the 11 – 20 messages
received category. This difference could again be accounted for by a greater empathy with
the sender that may be present as a result of academic disciplines.
The average received load for each group is as a follows, HCSILS received an average of
17.45 messages daily. Humanities received an average of 21.17 messages daily. ITES
received an average of 21.77 messages daily. SEC received an average of 21.60 messages
daily. Unspecified Academics received an average of 20.73 messages daily. All of the
groups are very similar except for the HCSILS group who received fewer messages daily.
Silverstone (2014) identified a mean received of 20.17 messages daily for the Academic
role.
As considered by Silverstone (2014), whilst high numbers of received messages are
considered to be an indicator of overload, alone the figures do not carry much meaning.
Taking sent and received load figures in conjunction with the perceived maximum number of
messages that can be sent and received will provide a context to assess overload in the
groups being analysed.
ANOVA tests show that there is no significant difference in the perceptions of sent message
manageability between the groups being analysed (p=0.474 in all cases). The unspecified
academic group was excluded for further analysis and there were still no significant
differences between the groups (p=0.469).
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Despite no significant differences between perceptions of sent message manageability,
there are differences to be observed between actual and perceived sent load. HCSILS,
unspecified and Humanities groups send fewer messages daily than are perceived to be
manageable with HCSILS sending levels lower than the perceived maximum than
Humanities. Unspecified lecturers send significantly less than they perceive to be
manageable. ITES and SEC groups send slightly more than the perceived manageable
maximum suggesting a degree of overload caused by sent message load.
When considering sent message load, ANOVA tests further demonstrated that there is no
significant difference in the perceptions of received message manageability between the
groups being analysed (p=0.384 in all cases). The unspecified academic group was
excluded for further analysis and there were still no significant differences between groups
(p=0.790).
Out of the groups being analysed, only the SEC group received fewer messages than they
perceived to be manageable. This suggests that the Academic role as a whole is overloaded
by received messages as identified by Silverstone (2014). Unspecified and ITES groups
receive an average of 9.02 and 9.37 more messages respectively than they perceived to be
manageable. Based upon the assertions of Ingham (2003) and Dabbish and Kraut (2006)
that received messages generate the greatest level of overload, it is suggested that these
two academic groups are suffering from the greatest email overload.
Perceptions of change help to further give context to the findings of sent and received
messages. Significant proportions of all groups felt that sent load had increased in recent
years. However, HCSILS and the Unspecified Academic group felt this most strongly. This is
contrary to the difference between actual and perceived maximums. Despite the descriptive
differences, statistical analysis via Chi-Square assessments fail to find any significant
differences between the groups (x=10.305, p=0.244).
Similarly, very high proportions of respondents felt that their received load had increased.
This was most strongly felt by the HCSILS group, matching the results for sent message
load increases. Humanities and Unspecified academic felt that received loads had increased
to a similar extent with ITES and SEC groups perceiving this increase to a lesser extent.
Despite differences being observed in the perceptions of changes in received message
load, these observations are not supported by statistical analysis via Chi-square testing
(x=9.124, p=0.332).
In order to compare the influence of micro cultures against the meso-cultures identified by
Silverstone (2014), the ranking exercises conducted looking at the differences between
actual and perceived maximums for sent and received messages against percentage
increase will be repeated. Figure 12 below illustrates this in the same format used by
Silverstone (2014).
The same pattern is not observed when analysing micro cultural influence in the Academic
role as observed by Silverstone (2014). However, it is clear that all academic groups are
receiving more messages than they perceive to be manageable. Dabbish and Kraut (2006)
identified that received message loads are a key influencing factor in email overload. The
results in figure 12 suggest that overload may exist for all groups analysed.
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Figure 12, the relationship between differences in actual load and perceived manageability related to
the perceptions of increase, results shown by Academic Employment Group
Mean
Manageable
15.88
14.49
16.49
13.14
13.05
Diff.
Rank
HCSILS
Humanities
ITES
SEC
Unspecified
Mean
Sent
11.83
13.66
14.39
9.79
12.96
-4.05
-0.83
-2.10
-3.35
-0.09
5
2
3
4
1
Mean
Manageable
15.66
15.72
18.40
16.77
13.96
Diff.
Rank
HCSILS
Humanities
ITES
SEC
Unspecified
Mean
Received
17.45
21.17
21.77
21.60
20.73
+1.79
+5.45
+3.37
+4.83
+6.77
5
2
4
3
1
%
increase
90.5%
83.5%
80.0%
74.6%
86.3%
Rank
%
increase
93.5%
89.5%
81.4%
83.1%
88.7%
Rank
1
3
5
4
2
1
2
5
4
3
A correlation test was performed to investigate the relationship between perceived
maximums for sent and received messages. This test was performed by Silverstone (2014)
and showed a significant correlation between the two. For the whole sample group a
correlation of r=0.461, n=413, p=0.000. Whilst the correlation figure is not high, there is
clearly a statistically significant outcome. The results suggest that all academic groups feel
that sent and received message loads should be similar. With the unspecified group
removed r=0.583, n=259, p=0.000.
There are slight differences to which the figures correlate for each of the groups. For the
HCSILS the correlation is stronger with r=0.876, n=56, p=0.000. For the humanities group
the correlation is slightly less strong with r=0.749, n=75, p=0.000. For the ITES group the
correlation is slightly less strong again with r=0.716, n=62, p=0.000. For the SEC group the
correlation is much weaker with r=0.315, n=66, p=0.011. Whilst this is still statistically
significant, with a 2-tailed test of less than 0.5, it is less so that the other groups. Finally, the
unspecified academic group had a correlation of r=0.314, n=154, p=0.000, again, weaker
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
than other academic groups. This illustrates that there are some slight differences between
different academic groups as to whether sent and received load should balance.
The average time spent using email was investigated using ANOVA testing and did not
differ between the groups being investigated (p=0.097 including the unspecified group and
p=0.066 excluding the unspecified group).
The time spent per message was calculated in the same way undertaken by Silverstone
(2014). The HCSILS spent an average of 1.73 minutes per message, Humanities spent 1.44
minutes per message, ITES spent 1.24 minutes per message, SEC spent 1.23 minutes per
message and the Unspecified Academic group spent an average of 1.47 minutes per
message. This compares to the overall average of 1.43 minutes per message for the
Academic role as a whole identified by Silverstone (2014).
There is no clear indication as to why the time spent per message should differ in such a
way. Silverstone (2014) suggested that increased time per message may be an indication of
greater complexity included within the messages.
Whilst it is pure conjecture, Humanities subjects tend to require the conveying of concepts
and ideas which may require greater explanation where ITES and SEC subjects tend to be
more technical and specific. It may be that the approach to the subjects also informs the
approach to message content and the time spent writing them. However, it is not possible to
tell from this study.
Silverstone (2014) investigated the relationship between desire to change and employment
role. It was suggested that there was some difference between roles when linked with
cumulative differences between perceived maximums and actual load. In this case, an initial
Chi-square test was used to assess the level of relationship and it was found that there was
no significant difference between the groups investigated (x=4.431 p=0.351).
When compared to cumulative differences between perceived maximums and actual load,
the relationship observed by Silverstone (2014) is not repeated. This further suggests that
the limit to which ‘group’ has an impact on email usage is to be found at the meso-cultural
level. This is shown in figure 13 below.
Figure 13, The relationship between the difference in actual and manageable loads
and the desire to change, results shown by Academic Employment Group
HCSILS
Humanities
ITES
SEC
Cumulative
Difference
(actual
vs.
manageable)
-2.26
+4.63
+1.27
+1.48
Rank
Desire
change
(yes)
to Rank
5
2
4
3
33.9%
44.2%
32.4%
33.8%
3
1
5
4
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Unspecified
+6.68
1
41.9%
2
These findings do not support those of Dabbish and Kraut (1996) as there is no descriptive
relationship between the amount of overload caused by incoming messages and the desire
to change. If incoming load causes the greatest overload for users it could be expected that
they may wish to change their usage. However, this is not observed.
Volume and content management as well as the desire to receive fewer unsolicited emails
were cited as the main reasons for wanting to change email usage. These both tend to be
related to received message load rather than sent. In all cases, actual received load does
exceed perceived maximums. These issues are clearly having a negative impact on users in
all of the groups investigated.
Whether or not users consider others before sending emails was investigated. There appear
to be some differences but these are not considered to be significant (x=8.876 p=0.070).
The significance figure is only slightly above the p=0.050 considered to be significant
demonstrating that there is some difference worth considering.
The Humanities group was most likely to consider others with the HCSILS group least likely
to do so. The Humanities group was most likely to consider the appearances and
interpretation of their messages as a reason for considering others than the other groups.
Where users in the SEC group considered others they very strongly identified with the need
to consider others time management, whether or not the message is actually needed and
the need to manage the expectation of others. The SEC group also has the lowest level of
sent messages. However, there are no other precedents to compare this with.
The high levels of consideration of others demonstrate that the concerns of Denning (1982),
Seeley and Hargreaves (2003), Ingham (2003) and Evans and Wright (2008) have been
taken on board and the generation of excessive load by not considering others appears to
be thought about. However, there is still load above that which is manageable and therefore
it has to be questioned as to whether users wish to be perceived as considering others more
than they actually do.
Perceptions of wastage differ between the groups (figure 7). The Humanities group is most
likely to perceive that time is wasted with 75.3%, higher than the overall average for the
Academic role at 63.1%. This is much higher than the ITES group where only 44.3% of
respondents feel that time is wasted. Chi-square testing reveals that the difference observed
between the groups is statistically significant (x=16.410 and p=0.003 including the
unspecified group and x=16.289 and x=0.001 excluding the unspecified group).
Silverstone (2014) noted that there may be a pattern between sent and received messages
and perceptions of wastage. However, this observation did not hold true to Academics
where a disproportionately high proportion felt that time was wasted when considered in the
context of sent and received messages. The same is observed here. There is little
difference between the average loads of the Humanities and ITES groups but there is a
difference in the perception of wastage.
As well as different proportions of each group feeling that time is wasted there is also
difference in the proportion of time wasted. Humanities felt that the greatest proportion of
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
time is wasted at 20.94%. Following the same pattern as with the proportion of those who
felt time is wasted, ITES reported a lower figure of overall wastage but it was not the lowest
at 17.58%. The SEC group reported 15.58% wastage. Whilst these figures are different,
crucially, ANOVA testing reveals that there is no significant difference between the groups in
terms of proportion of time wasted (p=0.422 including the unspecified group and p=0.319
excluding the unspecified group).
Where groups identified strongly that time is wasted the most common reason identified is
work related emails that are not relevant or are duplicated. The proportion of respondents
identifying this follows the same pattern as those who identify that they waste time in the first
place. This suggests that actual load has less of an impact on academic groups than the
issue of duplication.
ITES staff, with the lowest proportion identifying that time was wasted also had the lowest
proportion identifying the drawback discussed. Humanities, with the highest proportion
reporting that time is wasted had the highest proportion identifying the drawback. HCSILS,
SEC and Unspecified who had very similar proportions identifying that time is wasted also
had similar proportions identifying the drawback. There were no other significant reasons for
wastage identified by the respondents.
There is no evidence to suggest why there should be so much difference in opinion within
the Academic role. Once again it may have something to do with the nature of the work
undertaken by the groups. The ITES group contains academics involved in delivering
computing and IT based subjects. As such it is fair to assume that they may have a greater
applied understanding of how to manipulate the technology to reduce the number of
unwanted messages. It may also be that the nature of work undertaken means that there
are fewer irrelevant messages as work culture may limit sending relevant content only.
If the data were gathered from one institution it may be possible to track the difference in
perception of irrelevant or duplicated messages back to a culture generated by management
style. However, with respondents from 20 institutions it is not possible to assume this.
Silverstone (2010) identified that irrelevant messages placed an excessive load on users.
With perception of waste as an indicator of overload it can be assumed that this
consideration can be applied in this case. Sumecki et al (2010), Whittaker and Sidner
(1996), Dabbish and Kraut (2006), Adam (2002) and Jackson et al (2003) have identified
irrelevant messages as an issue in email usage. Interestingly, this issue was not identified
as a drawback to email when respondents were questioned about the drawbacks and
benefits. This suggests that perhaps it is not consciously considered.
The vast majority of respondents had not attended training in the twelve months up to the
questionnaire deployment. Of those that did the relevance to the group was varied. SEC
found the training least useful with only 35% believing it to be of use. 63.2% of the HCSILS
group found training to be of use. The reported training focused on software or hardware
training. Very few respondents attended content management training.
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Figure 14 – statistical significance comparisons
Test
Silverstone (2014)
This study
Sent message load
X=235.516, p=0.000 With UA X=15.235,
(Chi Square two
p=0.229
tailed, significance
Without UA x=912,
p=0.05)
p=0.179
Received message X=237.404, p=0.000 With UA X=8.912,
load
p=0.179
(Chi Square two
Without
UA
tailed, significance
x=14.014, p=0.122
p=0.05)
Sent
message Whole
study With UA, whole
manageability
(p=0.000) A & all study (p=0.474)
(one way ANOVA (p=0.000) MM & BS Without UA, whole
test, significance at (p=0.681) MM & SM study (p=0.469)
p=0.05)
(p=0.012) BS & SM
(p=0.001)
Received message A & all (p=0.000) With UA, whole
manageability
MM & BS (p=0.149) study (p=0.384)
(one way ANOVA MM &SM (p=0.022) Without UA, whole
test, significance at BS
and
SM study (p=0.790)
p=0.05)
(p=0.000)
Increase in sent load X=15.149, p=0.19
(x=10.305, p=0.244)
(Chi Square two
tailed, significance
p=0.05)
Increase in received X=10.043, p=0.123
X=9.124, p=0.332)
load
(Chi Square two
tailed, significance
p=0.05)
Correlation
for R=0.736,
n=848 With UA r=0.461,
perceived maximum P=0.000
n=413, p=0.000
sent and received
Without UA r=0.583,
(bivariate Pearson’s
n=259, p=0.000
correlation,
two
tailed, significance
at p=0.05)
Time spent using Whole
study With UA, whole
email
(p=0.000) SM & BS study (p=0.097)
(one way ANOVA (p=0.000) SM & A Without UA, whole
test, significance at (p=0.000) SM & A study (P=0.066)
p=0.05)
(p=0.997) BS + A &
all (p=0.000)
Desire to change X=55.141, p=0.000
X=4.431, p=0.351
usage (Chi Square
two
tailed,
significance P=0.05
Consideration
of X=3.926, p=0.270
X=8.876, p=0.070
others when sending
email (Chi Square
two
tailed,
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
significance P=0.05)
Perceptions
of
wastage (Chi square
two
tailed,
significance P=0.05)
Amount of wasted
time
(one
way
ANOVA
test,
significance
at
p=0.05)
Number of items
where significance is
achieved
X=31.792, p=0.000
With UA x=16.410.
p=0.003
Without
UA
x=16.289, p=0.001
Whole
study With UA p=0.422
p=0.016
range Without UA p=0.319
(p=0.008 to 0.981)
9
2
5. Conclusions
Some differences in the way that the groups identified view and use email have been
shown. However, there is little evidence to suggest that these differences are statistically
significant. Sent and received loads, the desire to change, perceptions of manageability,
average time spent and consideration of others did not differ statistically. This suggests that
micro cultures do not have an influence on email usage, perceptions and behaviour.
The relationships shown by Silverstone (2014) at the meso-cultural level in terms of the
relationship between actual and perceived load and perception of increase do not persist
when looking at groups at this level. Similarly the relationship between actual and perceived
load and the desire to change does not produce the same pattern at the academic micro
cultural when compared to the meso-cultural level.
The only significant differences observed were seen in the perceptions of wastage which is
similar to the findings of Silverstone (2014). In addition, the correlation between perceived
manageability of sent and received messages persists in this analysis. Only 2 of 12
statistical markers are significant which compares to 9 of 12 observed by Silverstone (2014).
Figure 14 in the discussion section illustrates the statistical significances observed by
Silverstone (2014) and those observed in this paper.
References
Adam, R 2002, ‘Is e-mail addictive?’ Aslib Proceedings, 54, 2, pp85 – 94
Collins, H 1986, ‘Electronic mail’, Facilities, 4, 12, pp10 – 14
Cunningham, H & Greene, B 2002, ‘Before you hit send – getting e-mail communication
right, why e-mail etiquette is a critical communication issue’, SCM, 6, 5, pp6 – 20
Dabbish, L & Kraut, R 2006, ‘Email overload at work: an analysis of the factors associated
with email strain’, CSCW, November 04-08 pp431 – 440
Denning, P, 1982 ‘Electronic junk’, Communications of the ACM, 25, 3, pp163 - 165
Evans, C & Wright, W 2008, ‘To all users:Copy all users’, Management Services, 52,1,
pp24–27.
Huczynski, A & Buchanan, D 2001, Organisational behaviour: an introductory text, Pearson,
UK
Proceedings of International Social Sciences and Business Research Conference
4 - 5 December 2014, Hotel Himalaya, Kathmandu, Nepal, ISBN: 978-1-922069-65-8
Ingham, J 2003, ‘E-mail overload in the UK workplace’, Aslib Proceedings, 55, 3, pp166 –
180
Jackson, T, Dawson, R & Wilson, D 2003, ‘Reducing the effect of email interruptions on
employees’, International Journal of Information Management, 23, pp22 - 65
Kierkegaard, S 2005, ‘Privacy in email communication, watch your email: your boss is
snooping’, Computer Law and Security Report, 21, pp 226 – 236
Miller, N 1999, ‘Email abuse and corporate policies’, Network Security, pp 13 – 17
Seeley, M & Hargreaves, C 2003, Managing in the email office, Butterworth-Heinemann, UK
Seshadri, S & Carstenson, L 2007, ‘The perils of e-mail communications in non-profits’,
Non-Profit Management and Leadership, 18, 1 pp77 – 99
Sillince, J, Macdonald, S, Lefang, B & Frost, B 1998, ‘Email adoption, diffusion, use and
impact within small firms, a survey of UK companies’, International Journal of
Information Management, 18, 4, pp23` - 242
Silverstone, B 2012, ‘Developing a relationship centred communication framework for email
selection and usage – a literature review’, World Journal of Social Sciences, 2, 7,
pp257 – 269
Silverstone, B 2014, ‘The influence of role on email usage profiles, a study of the welsh
further education sector’, The Macrotheme Review, 3, 2 pp16 – 37
St Amant, K 2001, ‘Success in the International Virtual Office’ in JOHNSON, N (Ed.),
Telecommuting and the Virtual Offices, Issues and Opportunities, Hershey, USA
Sumecki, D, Chipulu, M & Ojiako, U 2010, ‘Email overload, exploring the moderating role of
the perception of email as a ‘business critical’ tool’, International Journal of Information
Management, pp1 - 8
Trice, H & Beyer, J 1984, ‘Studying organisational cultures through rites and ceremonies’,
Academy of Management Review, 9, 4, pp 633 – 669
Whittaker, S & Sidner, C 1996, ‘Email overload: exploring personal information management
of e-mail’, CHI, vol. 96 April 13 – 18