After building regular HVMs from the interviews, the study tried several methods by collating and arranging data from interviews to attain a proposed HVM that can help designers better determine target users’ attitudes and needs, and carry out the corresponding design specification in this new area. The final suggested method is presented as follows.
4.2.1 Sorting consequences
As a result of numerous consequences mentioned in the interviews, it’s often to see that participants used different consequences to describe a same feeling due to their personal idioms, and it would gain complexity and confusion in building HVM.
Thus, the study sorted these consequences by their representing meanings, and picked a consequence as the representative of the group. Although Walker and Olson suggested the six levels in the means-end chain (see Table 2-2), the consequences attained from the interviews were not distinct enough to classify into six layers. Thus, we took their suggestion to classify all the consequences into two categories, on one hand, the functional consequences, illustrate the participants’ feelings, affected by the attributes of services and applications. On the other hand, the emotional consequences, affected by functional consequences and be in response to participants’ personal life experience.
Meanwhile, there were several negative feelings described during the interviews. We extracted these negative consequences, and did the sort and
categorization too. In the end, the study would get a group of positive consequences and the other group of negative consequences, with both sorted.
Consequently, linking attributes, functional consequences, emotional
36
consequences and human values in a hierarchical order to rebuild the HVMs of each participants according to the results of interview (see Figure 4-5).
Figure 4-5 the HVMs of three subjects
4.2.2 Crossing the HVMs
On the previous stage, we got the HVMs in several contexts. According to the main purpose, the study wants to investigate values in the wellness and the feelings related to design factors from APPs. However, on the HVM of wellness services, the attributes are far from the context of applications, but the values linked from them represented the attitudes toward health issues. Hence, the study disregarded its attributes and functional consequences, but kept its emotional consequences and the linkages to values, which showed what participants pursue in the wellness.
On the HVM of other applications, the values are not acquired in the context of wellness, which are not the subject of the study, but the related consequences are derived from the attributes of APPs that we want to apply as design specification.
Thus, we abandoned the linkages between emotional consequences and values, but
37
reserve the linkages between functional consequences and emotional consequences.
It is notable that the two wellness applications, Nike+ and Fit 4 Rhythm, are in the context of wellness and associated with applications. Thence, the study kept the functional consequences, emotional consequences, values and the linkages between them (see fig 4-6).
Lastly, we got different parts from these HVMs, which show context in the wellness or applications, but in a fragmented structure. The study wants to deliver a proposed HVM that can present not only wellness values, but also feelings associated with APPs in an integral framework that can easily understood and employed by designers. Hence, we assumed the common emotional consequences that showed up in all three HVMs as Key Nodes to consolidate the extracted parts from these HVMs and came out the proposed HVM, P-HVM.
Figure 4-6 crossing the HVMs to attain P-HVM
The structure of P-HVM is shown in Fig 4-7, the linkages route from top to the lower layer (grey arrows), in addition, there are some links point to the elements in the same layer (grey loop arrows) before they point to the lower layer, which
Attributes of
Wellness HVM APPs HVM Wellness APPs
HVM P-HVM
38
describes the users’ mental context in details. It’s notable that we won’t adopt attributes from any HVMs; on the contrary, we kept it for designers to employ corresponding design factors on the basis of lower levels in the P-HVM.
Fig 4-7 the structure of P-HVM
4.2.3 Weighting the P-HVM
After the consolidation of the P-HVM, the study found that several linkages point to the same nodes would confuse designers that which route is the more influential one links to the target value. Thus, we sent out questionnaire to bring out the weight of each linkage to help designers make trade-offs when they are struggled with design. For example:
Recorded (FC) -> Contextual (EC) -> Progress Practical (FC) -> Accompanied (EC) -> Progress
These two ladders pointed to the same value, Progress, and gave rise to the consideration that which is more influential for pursuing progress. The two questions were delivered to get the answer:
I think the wellness service involves in life context will help me in the pursuit of progress.
39
I think the wellness service accompanied by friends will help me in the pursuit of progress.
There were total 139 questions in the questionnaire, each of them represent a linkage of multi-choices in the P-HVM like this. 83 participants selected the grading from “strongly agree”, weights 5 points, to “strongly disagree”, weights 1 point, according to their personal experiences and feelings toward the questions. The number of people chose the option multiply its weight would get the weight of the linkage.
In order to keep the P-HVM easy to interpret, the study calculated the average of the weights that point to the same node, and stressed the linkages that is above it, which means they are more influential to attain the end values. For example, the emotional consequence “Sharing” is pointed by the functional consequences “Fun”
with the weight 307, “Various” with the weight 302, “Additional” with the weight 300,
“Cost-effective” with the weight 328 and “Convenience” with the weight 317. The average of these five is 311, thus it’s notable that “Cost-effective” and “Convenient”
are above it and more important than others (see Table 4-1).
EC\ FC Fun Various Additional Cost-effective Convenient
Sharing 307 302 300 328 317
Table 4-1 the functional consequences related to “Sharing”
The emotional consequences that point to the value “Progress” are shown in Table 4-2. “Inspired” weights 349, “Fulfilled” weights 365 and “Corresponding”
weights 350 are above the average 342. Thus, we can consider them first to attain the value “Progress”.
40
V \ EC Accompanied Inspired Fulfilled Corresponding Contextual
Progress 319 349 365 350 326
Table 4-2 the emotional consequences related to “Progress”
4.2.4 The P-HVM in the wellness
After consolidating all the consequences and values, and receive the weights on the linkages, we now can build up the consolidated P-HVM from the interviews. The P-HVM of value “Quality” extracted and shown in Fig 4-8, the darker and thicker arrows are the ones with the weights above average, which are more influential to attain the value. For example, the emotional consequence “Attractive” is more important than “Effective” when facing the value “Quality”. After routing from
“Attractive”, we can be told that “Various”, “Fun” and “Contextual” are the suggested design specification for it rather than “Out of the ordinary” and “Occasional”.
41
Fig 4-8 the HVM of value “Quality”
The P-HVM of value “Effectiveness” is shown in Fig 4-9. The route “Recorded ->
Real -> Fulfilled”, “Practical -> Fulfilled” and “Various -> Fun -> Fulfilled” are more influential to get the target value “Effectiveness”, which can help designers make trade-offs when delivering a related design. There are total 17 P-HVM of the values in the wellness, shown in Appendix C.
42
Fig 4-9 the HVM of value “Effectiveness”