What Explains Differences in Users’ Inclination to
Appropriate Technology for Unconventional Purposes?
A Preliminary Analysis
Antti Salovaaraa, Sacha Helfensteinb, Mikael Wahlströma, Antti Oulasvirtaa
a Helsinki Institute for Information technology
Helsinki University of Technology and University of Helsinki
P.O.Box 9800, 02015 TKK, Finland
b Agora Human Technology Center
University of Jyväskylä
P.O. Box 35, 40014 Jyväskylä, Finland
It is common to state that inventions of new purposes of use
arise in social interaction with other technology users. Social
aspects of appropriation have subsequently received a lot more
attention than individual users’ characteristics in appropriation
research. To remedy this imbalance, this paper presents a pre-
liminary analysis of a web survey that charted aspects of digital
camera use and individuals’ photography orientations and used
them as predictors of digital camera appropriation. Gender,
technology understanding and exchange of ideas with others
proved tentatively the best predictors of appropriation.
Appropriation, web survey, user characteristics.
ACM Classification Keywords
H.1.2 [Models And Principles] User/Machine Systems – hu-
During the past two decades, it has been increasingly recog-
nized that information systems, like any tools, are not only used
for purposes specifically designed for them, but that they are
appropriated for purposes that can surprise their designers. The
inventions of new purposes of use are called appropriations.
The reasons for appropriations are both social and individual.
The social processes contributing to appropriations have al-
ready attracted enthustiastic attention in the research commu-
nity. The question who decides how technology should be used
(e.g., at a workplace) has been addressed in science and tech-
nology studies  and management studies . Another often-
studied topic has been how appropriations are promoted and
adopted by certain people at a workplace [3, 4].
However, an individual user’s role in the invention process has
been a lot less studied topic. In particular, research on the role
of cognitive processes and individual user’s characteristics has
been almost non-existent. In the few studies available it has
been found that users with a learning-centered work orientation
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies
are not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
ECCE 2009, September 30 – October 2, 2009, Helsinki, Finland.
Copyright 2009 ACM 1-58113-000-0/00/0004…$5.00.
have been found more willing to learn new uses of technology
even when it can be hard, than users with a performance-
centered orientation who are willing to learn only if learning is
easy . In addition to the learning vs. performance orientation,
in a mobile phone context, evidence has been found that users
can also be segmented into different kinds of adopters whose
inventions of use are different . Cyclical perception-action
models of appropriation processes have also been presented .
However, these studies have ignored the question what individ-
ual characteristics are important for appropriation. This paper
addresses this question by focusing on one easily appropriable
technology – digital cameras. In addition to compact and single-
lens reflex (SLR) cameras, also mobile phones are nowadays
often equipped with cameras that have a reasonable picture
quality. The versatility to take photos with these technologies
has provided a fruitful basis for ingenious uses, eight of which
have been used as a measure of appropriation in this paper. The
goal of this study was to identify user characteristics that could
statistically predict such uses.
2. RESEARCH APPROACH AND MODEL
The research approach described in this paper is different from
many of the previous studies on appropriation. Compared to the
sociologically oriented studies in which the focus has been on
charting the complexity of the phenomenon and aiming at a
holistic picture, the approach here is focused, narrowly defined
and aims for measurable results. Appropriation is here inter-
preted as an invention of a new purpose of use, previously un-
known to the user.
2.1 Antecedent Factors of Appropriation
No theory has been presented that would attempt to list the
possible antecedent individual user’s characteristics that con-
tribute to appropriation. Therefore, for the purpose of this study,
a set of tentative antecedent constructs related to digital camera
use were generated by researchers. These were:
• Setting up personal Goals for one’s photography activities
(e.g., personal projects or directions of improvement).
• Reflection of one’s practices by evaluating one’s shots.
• Having a comprehensive and correct Mental Model of how a
camera works and what its functions are.
• Curiosity of trying new ways of photography.
• Taking photos spontaneously and in Ad Hoc ways, in a spur
of action, without always thinking before acting.
• Having a broad understanding of the surrounding technology
Ecology; e.g. how photos can be edited or used in other me-
• Awareness that a digital camera is an easily appropriable tool
and thus a potentially useful in many situations.
• Social Construction: learning new ways of use from others
through teaching, observation or exchange of ideas. The pur-
pose of this construct was to evaluate the importance of some
of the previously studied factors of appropriation.
Because the constructs were not drawn from existing theories,
the nature of this study was exploratory and proto-theoretical. It
was also conceptually organized in a top-down manner. Each
construct was divided into sub-constructs, and phrased as a
statement that could be answered on a Likert scale (1=totally
disagree, 5=totally agree). For instance the Mental Model con-
struct was represented of the following sub-constructs:
• Learning camera’s functions comprehensively (”I have fa-
miliarized myself with more or less all the functions in my
camera or cameras”.
• Understanding cause and effect (”I know how to tune the
camera settings to capture photos with good quality”).
• Knowing the good and bad aspects of one’s cameras (“I
know which are the most important strengths and weak-
nesses of my camera or cameras”).
All in all, 35 Likert scale statements were generated to repre-
sent the eight constructs. The constructs and their wordings
were improved iteratively by organizing two pilot studies. On
the other hand, it was admitted that the set of constructs could
not be complete, leaving a possibility that another set of con-
structs could be a better predictor of appropriation.
2.2 Measuring Appropriation
To evaluate the tentatively postulated antecedent constructs, a
measure was needed for appropriation. To start with, the fol-
lowing eight uses were defined as signs of appropriation:
1. Mirror: pointing the camera toward oneself, in order to see
e.g. how one’s face looks like.
2. Map: taking a photo of a map, and using that photo as a re-
placement of a paper map.
3. Note-taking device: using the camera for note-taking when
the content is very visual, e.g. when shopping clothes.
4. Scanner: storing printouts and texts as images with a camera.
5. Memory card: plugging the camera into a computer like an
USB memory stick (does not work with all models equally).
6. Lamp: exploiting the camera as a light source.
7. Instructing device: using a sequence of photos to provide
8. Periscope: inspecting places that are otherwise inaccessible
to human vision.
This list of uses was also researcher-generated. During the two
pilot studies, respondents were asked to provide their own sug-
gestions for unconventional uses. However, the suggestions
received could be subsumed in already listed uses, or their
meanings were conceptually unclear. In the final questionnaire,
the set of uses was fixed to the eight uses listed above, to gather
a homogenous dataset of answers.
The questions in the web survey about each appropriation were
tree-structured. The starting question addressed the respon-
dent’s familiarity of using any digital camera in the given way
(e.g., as a mirror). The scale ranged from 1=“this use has never
occurred to me before” and 2=“I have know that this is possi-
ble, but I have never done so” to 6=“This is one of my estab-
lished uses for a digital camera”. Later questions addressed the
accuracy of the memory of the first time when the use was
learned, and the actual person who did the invention in the
From this eightfold set of tree-structured answers the measure
for individual’s appropriation was calculated using the familiar-
ity variable. For each of the eight uses, two new dichotomized
variables were created, one having value “0” if the respondent
replied with an answer coded as 1 (see the values above), and
“1” otherwise, the other having a value “0” if the respondent
replied with an answer coded as 1–2 , and “1” otherwise. Thus,
one variable expressed whether a user was familiar with an
appropriation or not, the other whether he or she had ever used
it or not. Summing these binary variables over the eight uses
yielded measures for the overall degree of appropriation (Total
Degree of Appropriation Familiarity and Total Degree of Ap-
propriation Employment), respectively, both ranging 0–8. The
use of two measures was deemed important since the purely
imagined use (i.e., answer coded as 2) was a common choice in
the data, gathering on average 21% of all the familiarity an-
swers. It was important to know whether its exclusion from the
appropriation degree would change the results considerably.
3. THE STUDY
The Likert scale statements about the antecedent constructs as
well as the tree-structured questions about different appropria-
tions were part of a web survey in Finland between November
2008 and January 2009. The survey contained also other items,
the most important ones from the point of this study being the
demographic details (gender, age, education among others),
camera use experience (expressed as years of film, digital and
phone camera use) and camera use frequency (asked separately
for each type of camera, ranging from daily to terminated use).
Each respondent was also asked to assess whether she or he
considered herself or himself as a beginner, novice, snapshot
taker, amateur, expert/professional, or other kind of actor in
photography. The questionnaire could be answered anony-
mously. Two pilot studies were organized before the actual
Due to the mundaneness of a digital camera as a consumer tech-
nology, reaching a high number of responses was deemed more
important than a strict probability sample of respondents. Invi-
tations to participate were distributed to authors’ social circles,
camera-related web forums, and camera clubs. By buying key-
word-based advertisements from a large Finnish search com-
pany, the survey was also visible at different pages in the web.
Respondents could also invite their friends by providing their
email addresses. These addresses were not archived in the data-
base. A raffle of fifteen 20 EUR gift coupons was arranged
between those respondents who had completed the survey.
The survey reached N=2388 of complete answers from digital
camera and/or phone camera users. The distribution of values in
the self-reported expertise variable shows that the recruitment
from camera clubs resulted in high number of answers from
amateurs: novice (8.7%), snapshot taker (36.2%), amateur
(47.9%), expert/professional (5.5%) and other (1.7%). Genders
were equally represented in the data (males 53.3%).
3.1 Are Appropriations Invented Alone?
Two questions precede an analysis of individual users’ charac-
teristics: 1) whether the eight unconventional uses are rare
enough to be informative of appropriation, and 2) whether they
are learned individually and not only from others). In light of
the data, the answer is clearly “yes” to both questions. Regard-
ing the first question, as visible in Figure 1, the variance of
values of Total Degree of Appropriation Familiarity is large,
meaning that people exhibit appropriation to different extents.
To answer the second question, on average 54% of the respon-
dents familiar with a use could remember the moment of learn-
ing “very well” or “partly”. Among these people, inventing the
use alone, without a help of others, was the most common con-
text of invention (39%), learning from others being the next
(21%). Appropriation by an individual is therefore common
enough to warrant a study of its antecedent factors.
3.2 Scope of Analysis in This Paper
In the following analysis, the focus has been limited to snapshot
takers only (N=. By this decision, the results are less susceptible
to a possibility that active photographing turns out to be the
underlying reason for appropriation Such a biased result is less
probable in a snapshot taker data. Due to this scope, the results
are preliminary and will be extended in future work.
The reliability and validity of the antecedent constructs were
evaluated for discriminant validity and internal consistency. For
discriminant validity, a confirmatory factor analysis for the
Likert statements was carried out. Based on the results, some of
the initial constructs were combined into larger ones, and
Awareness was dropped because of poor factor loadings. These
changes yielded the following constructs for the actual analysis:
• Setting up personal Goals (unchanged).
• Social construction (one item dropped).
• Technological knowledge (Mental Model and Technology
• Exploration and learning: (Curiosity with parts of Reflection
and Ad Hoc photography style added): Users scoring high on
this dimension display increased motivation to actively dis-
cover best use practices and learn from experiences.
• Self-concept (items in Reflection construct related to one’s
tendency to analyze oneself as a photographer): Users scor-
ing high on this dimension display an increased sense of self-
critizism regarding their photography and camera use skills.
To evaluate internal consistency of the constructs, Cronbach’s
alpha coefficients were calculated, attaining values between .76
(goals) and .89 (technological knowledge), which was inter-
preted as a strong support for the synthesized scales
4.1 How Appropriators and Non-
For each of the eight appropriations, analysis was carried out to
see how users that (a) were familiar with an appropriation (i.e.,
whose dichotomized familiarity value was 1 or above) or (b)
had ever employed an appropriation (i.e., familiarity value was
2 or above) differed from users that were not familiar (i.e.,
whose value was 0) or had never employed an appropriation
(i.e., values 0-1), respectively. The intention was to reveal those
basic user characteristics that can potentially distinguish appro-
priators from non-appropriators. U-tests, t-tests, and Chi-Square
tests were used to assess the differences between those groups.
The analysis provided an overall proof that appropriators can be
set apart from non-appropriators in basically all construct di-
mensions. In addition, people that were familiar with the eight
uses also displayed differences regarding age, gender, as well as
camera use history duration and use frequencies. In general,
appropriators (i.e., those familiar with and using appropriations)
had higher mean scores in all constructs, except Self-concept,
which was associated to appropriation only for the Instructing
device, Periscope, and Memory card use purposes.
Appropriation, as measured by familiarity level, was generally
more common among men, except for the Mirror and the Lamp
appropriations, both of which seem more general purpose or
more valuable to women as well. Also, familiarity was on aver-
age more wide-spread among younger users.; however, not
always in terms of the actual employment into use. In fact, us-
ers familiar with the Instructor appropriation proved on average
older than others. Finally, appropriators seemed to be on aver-
age more active digital and phone camera users, and also had
longer experiences with using digital camera devices.
4.2 Basic User Characteristics and
Constructs as Predictors of Appropriation
In order to test for a statistical significance of user characteris-
tics as predictors of the total degrees of appropriation (regarding
both familiarity and employment), hierarchical, stepwise multi-
variate regression analyses were used. The purpose was to iden-
tify user characteristics that can well explain the overall vari-
ability in appropriation among users.
In a first block, the fundamental person variables (gender and
age) were entered as regressors, followed by the five improved
constructs (i.e., Goals, Social construction, Technological
knowledge, Exploration and learning, and Self-concept), and
finally by the remaining basic user characteristics (durations
and frequencies of use for film, digital, and phone cameras).
The research interest was to find out 1) which predictors are
yielded as significant, and 2) how substantially each of the three
regressor blocks increases the strength of the prediction model.
Since the analysis presented in the previous section had shown
that basically all user characteristics were related to one or an-
other appropriation, a decision was made to enter all variables
Figure 1. Distribution of how many of unconventional uses
the respondents were familiar with, plotted separately by the
different photography expertise levels.
into the initial regression model specification. For the Total Download full-text
Degree of Appropriation Familiarity, SPSS developed through
hierarchical stepwise regression a prediction model in five
steps. The resulting predictors were given as gender, age, Tech-
nological knowledge, Social construction, and frequency of
phone camera use. Calculating the regression with these predic-
tors only, yielded a model with R2 = .24, F(5,691) = 44.02, p <
.001, with gender (b = .23, t(691) = 6.49 , p < .001), and Repre-
sentation (b = .22, t(691) = 5.79, p < .001) the greatest predic-
Although the prediction model improved significantly (statisti-
cally speaking) with the addition of each of the three groups of
user characteristics (i.e., person variables, constructs, use pa-
rameters), Technological knowledge and Social construction
constructs (R2 Change = .11, F(2,692) = 47.78, p < .001) had a
substantial and clearly stronger effect on advancing the model
strength than for instance use frequency (R2 Change = .02,
F(1,691) = 19.14, p < .001). And, when compared to gender and
age in separate regression models, Technological knowledge R2
= .18) and Social construction (R2 = .11) proved better predic-
tors of users’ Total Degree of Appropriation Familiarity.
For the Total Degree of Appropriation Employment, on the
other hand, the regression analysis yielded a prediction model
including gender, representation, social construction, phone
camera use frequency, and duration of digital camera use his-
tory. The resulting model is described with R2 = .26, F(5,671) =
47.65, p = .000, with representation as the strongest predictor, b
= .24, t(671) = 6.42, p < .001. Again, a substantial predictive
relevance was attested for the construct scales of Technological
knowledge and social construction. In fact, the regression
model strength appeared primarily attributable to these two
constructs, R2 Change = .14, F(2,673) = 58.82, p < .001, when
compared to the model including gender, R2 Change = .08, and
use frequency and duration history, R2 Change = .05.
The regression analyses point towards a tentative conclusion
that Technological knowledge and Social construction are im-
portant factors contributing to appropriation, at least in a digital
camera use context. That is, to generalize, acquiring a good
understanding of how digital devices work, which functions
they have, and on the other hand, exchanging ideas about their
use with others, seem to be aspects that should be particularly
supported when attempting to design easily appropriable tech-
Many questions remain for future work, however. Why Explo-
ration and learning was not rendered as significant predictor
warrants future research, as well as whether the findings hold
also among the more experienced users such as those who con-
sidered themselves amateurs or professionals.
AS acknowledges financial support from UCIT Graduate
School and Academy of Finland. SH acknowledges financial
support from SoTech Platform project funded by the Technol-
ogy Industries of Finland Centennial Foundation. The authors
Lingyi Ma, Erkki Komulainen, and Andreas Dünser for ideas.
 Eglash, R. 2004. Appropriating technology: an introduc-
tion. In Appropriating Technology: Vernacular Science
and Social Power, R. Eglash, J.L. Croissant, G. Di Chiro,
R. Fouché, Eds. University of Minneapolis Press, Minnea-
polis, MN, vii-xxi.
 Loraas, T. and Diaz, M.C. 2009. Learning new uses of
technology: situational goal orientation matters. Int J Hum-
Comput St 67, 1, 50-61.
 Mackay, W.E. Patterns of sharing customizable software.
In Proc. CSCW’90. ACM Press, New York, NY, 209-221.
 MacLean, A., Carter, K., Lövstrand, L. and Moran, T.
User-tailorable systems: pressing the issues with buttons.
In Proc. CHI’90. ACM Press, New York, NY, 175–182.
 Orlikowski, W.J. 1992. The duality of technology: rethink-
ing the concept of technology in organizations. Organ Sci
3, 2, 398–427.
 Salovaara, A. 2008. Inventing new uses for tools: a cogni-
tive foundation for studies on appropriation. Human Tech-
nology 4, 2, 209-228.
 Wirth, W., von Pape, T. and Karnowski, V. 2008. An
integrative model of mobile phone appropriation. J Com-
put-Mediat Comm 13, 3, 593-61