Grant, Campbell, Chen, Cottone, Lapedis, and Lee (2007) showed that employees are willing to maintain their motivation when their work is relationally designed to provide opportunities for respectful contact with the beneficiaries of their efforts.
Our hypothesis is: Interdepartmental groups will more effectively collaborate when they are exposed to common beneficiaries.
If confirmed, our hypothesis could be an important development for Grant etal.’s research and contribute to the literature. Many industries in which there is high specialization in departments and for which client contact is not necessarily the norm may see an increase in collaboration. (Some examples include: investment banks; hospitals/healthcare clinics; and hotels/restaurants).
We think this could be especially relevant in the Health System (both the effect of the beneficiary contact and the import of the research) given human lives are the beneficiaries. Grant etal. (2007) showed this through their confirmation of Hypothesis 3: We predicted that task significance moderates the effect of contact with beneficiaries on persistence. When individuals are working on significant, high-impact tasks, they are aware that their efforts have the potential to benefit or harm other people. Accordingly, they are likely to invest additional time and energy in their efforts in order to benefit these people.
Data collection A, Positive/Reflexive (TBD): Ryan mentioned research showed that groups in which all departments were exposed to clients showed more willingness to collaborate interdepartmentally than when all departments had not been exposed to clients. This data will be looked at more carefully.
Data collection B, Reflexive: Andrew’s research with the Batten Institute contains data from at least one of the entrepreneurs studied that he was able to successfully improve interdepartmental performance by sharing feedback from a disappointed customer. The customer had ordered M&M’s with a personal message for a son’s birthday, and the order had arrived a day late. The mother expressed how upset her son was that his birthday was negatively effected by the tardy delivery. The entrepreneur then realized that his company did not merely manufacture personalized M&M’s, but something more important- happiness. He shared the story of this mother and her son with employees, and performance improved. This indicates there is probably value in looking back at the Batten data to see if other analogous stories are already collected. At the very least it indicates the potential to reexamine this entrepreneur’s organization and perhaps to interview other employees to see if they felt that the beneficiary exposure helped collaboration.
Data collection C, Positive: In order to test the hypothesis using positive methods, we ideally would propose recreating the experiment used to test Hypothesis 3 in Grant etal (2007) with a shift in emphasis from individual performance to group performance. Given the complexity of that experiment, however, perhaps we could use the Health System as research subjects and collect enough data to draw conclusions. We propose arranging meetings for representatives from all the relevant departments (to be determined when we have the full scope of the UVAHS organization) with patients. (Through prior healthcare experience, the allocation of patients to non-physician and nurse employees is through alphabetical assignment). It may be unethical to use a control group which is not exposed to patients. Thus, we hope to be able to get a sense of the control through interviews with managers who will be able to evaluate pre and post performance. (Generally healthcare facilities track data closely, and we may already have the control group benchmarks established, again this will be established when we have more of the data for the project).
If proven correct, one way to improve interdepartmental collaboration in the UVA Health System will be to expose representatives from each department to specific medical cases going forward.
While I think your data point descriptions are interesting and would yield fascinating findings, I think you guys need to do some work with the framing. Your first hypothesis, for instance, doesn't logically follow from the preceding paragraph. Grant studied individuals, not groups, and not between groups relationships. I think there is a logical argument to be made for extending Grant's analysis to these different settings, but you haven't made that argument yet. Separate note - I'm not sure I understand the term "reflexivity" as you're using it. Are you referring to it the way sociologists would to describe the observer-informant relationship?
ReplyDeleteI'm unclear on how you would test this? The M&M story is really interesting.. but how do you know it was that a group was motivated or just seperate individual's that got motivated by the story? I only ask this because we struggled with the same things..
ReplyDeleteIf their current processes do not expose certain team members to patients, I’m not sure how it would be seen as unethical to have a control group merely continuing their current standard operating procedures. It would also make the findings more measurable and likely compelling.
ReplyDeleteI think looking at beneficiaries this way is an exciting proposition. Just to add a note to that, I think the percieved coherence (by the inter-departmental actors) of the beneficiaries is important.
ReplyDeleteI too wonder about the M&M story. It seems clear that the entrepreneur was motivated by the story - but he had personal contact with the disgruntled customer. Perhaps the employees were not motivated by the story, but by their leader's direction to improve and his passion (after his improved motivation). Lots of possibilities and interaction effects to study here.
ReplyDeleteI like where you're headed with the idea; be sure to do a quick lit search to see if something similar hasn't already been done. I only say that because your next step is a very logical step given the Grant et al. (2007) piece.
ReplyDeleteI'm sure you haven't thought this far ahead yet but for what it's worth, be REALLY careful about what you choose as your DV here. No no, I don't mean patients haha (yes, that would be tough to get through the IRB). I mean what is your collaboration measure?
Remember your hypothesis is about collaboration, not performance. It might be tempting to see an increase in a performance measure (operationalized as manager interviews briefly in your write-up) and take that to mean that the group has collaborated more effectively since the performance measure increased.
This would not necessarily be the case. The individuals in the groups might just be working harder and not necessarily collaborating any better. In fact, the opposite of your first hypothesis might even be the case (increased competitiveness resulting from an increase in motivation and perhaps a motivation to take more of the credit for the work for the beneficiary).
But given what you say Ryan said, it sounds unlikely that the opposite would be the case, but your method should rule that one out.
I'm trying to think of ways you could creatively measure collaboration in this context and I'm sort of stumped. For one thing, it will likely be a fairly low power study once again (as most studies involving group DVs are). This suggests an added difficulty in that you're going to need to collect data from more people for each data point.
This probably rules out any form of observation measures on our part.
I've already given reasons why certain performance measures may not get at your true DV.
And that leaves something unfortunate like a pre-test post-test survey design...which would be simple enough I guess.
Let's think more about this, it might be supremely tricky to get at collaboration when there's bound to be plenty of performance data to look at. Maybe we could even tease out some of the collaboration construct by looking at particular types of performance data. Who knows...
Great thoughts all around. The M&M story is interesting, and indicates that there is probably merit in looking at that data, but it doesn't really tell us anything by itself. As for the hypothesis, Megan's right... I think another paragraph about the data ryan talked about could be enough to get us to the hypothesis. Frankly, there probably isn't any reason to state the hypothesis so formally after one setup paragraph. The DV is tough and the "collaboration" measure is tough, but from what I have seen with healthcare systems professionally is that time is one of the most tracked measures - in every way you can imagine. time to bill, time in the bed, time from consult to surgery, etc. I think the way to go would be to see what the current "issues" are (we don't know yet) see what they are tracking now and then craft something around that measure. Again, we will have to wait to see what we get from the UVAHS. Thanks.
ReplyDeleteGood post and good comments, once again. I like what you all are doing here!
ReplyDeleteI only have a couple of things to add to these already good comments:
1. It would be good to distinguish between theoretical and practical contributions. Usually, when academics say "contribution" they are referring to theoretical ones, but you seem to be referring to practical ones. That is fine, but you should be clear on audiences and what they expect when the time comes.
2. Building on Kelly/Sara's point, I find the M&M story intriguing for just the reasons that she is skeptical. As I mentioned in class, Adam had different effects when he tried to use a leader talking about beneficiaries as a proxy for beneficiaries. I don't remember the details of the study, but I think the M&M story contradicts his findings, at least partially, so it could open up a very interesting study about when and why leaders succeed at motivating employees.
3. I'm not doubting you, but I don't remember talking about research that says "groups in which all departments were exposed to clients showed more willingness to collaborate interdepartmentally than when all departments had not been exposed to clients." Can you give me a little more detail so I can remember what I said?