When problem solving prevents organizational learning

This paper from Anita Tucker & Amy Edmondson explored how problem-solving behaviours may focus too much on overcoming immediate obstacles that they inhibit broader organisational learning.

That is, people are too good at adapting and workarounds to make broken or inefficient systems work, and by doing so, obscure system improvement.

Providing context, they argue that:

·????????Short-term success of immediate problem solving “diminishes motivation to remove underlying causes of problems and because valuable data that can be used to justify and inform removal efforts are lost” and these behaviours reduce an organisation’s ability to detect underlying issues of recurring problems

·????????Problem solving can be divided into two types: first order and second order problem solving. First-order “allows work to continue but does nothing to prevent a similar problem from occurring”; workers exhibit first order when they don’t expend any more energy on a problem after working around the obstruction

·????????Second-order problem solving involves altering underlying causes to prevent a recurrence. First and second order problem solving are analogous to Argyris and Schon’s single-loop versus double-loop learning

·????????Prior research on problem solving focused heavily on identifying preferable methods rather than on what happens when people confront challenges and how they react; thus, this study sought to study people in their normal context

The researchers conducted 197 hours observing the work of front-line workers (hospital nurses), whose jobs involve many problem-solving opportunities. A goal from the researchers was to unearth the heuristics that workers use to solve problems that also unintendedly hinder broader organisational learning and then propose cognitive, social and organisational factors that reinforce the heuristics.

Results

From 197 hours of observations of 22 nurses, the researchers documented 120 problems; or approx. one problem every 1.6 hours of observation.

Kinds of problems nurses encountered included:

·????????missing or incorrect information;

·????????missing or broken equipment;

·????????waiting for a resource; and

·????????missing or incorrect medication

A key finding was that first-order problem solving prevailed as the most dominant heuristic. They noted that nurses “appeared virtually unable to engage in the kinds of second-order problem-solving behaviors prescribed in the literature” (p127).

Rather, first-order problem solving, which allowed work to continue and thereby ensuring immediate patient care also “ignored the possibility of investigating or changing causes of the problem” (p127).

Further, nurses were found to refrain from sharing information that could have been used to improve learning across the organisation.

Second-order problem solving was only used for a “handful of the problems” and when it was used, it was “was usually only in a minimal fashion, such as using an opportune moment to relay information” (p127).

Problem solving heuristics

Two key heuristics were observed and this characterised the response for 92% of issues:

1)?????do what it takes to continue the care of the patient; and

2) involve people with whom you are most comfortable, rather than the ones who are best able to solve the problem

The first heuristic, doing only what it takes to continue patient-care, was characterised by concern for securing the info or materials the nurse needed to continue caring for patients and not on understanding what caused the problem to occur. After the immediate problem was resolved then nurse didn’t expend any further energy on trying to investigate or solve the issue, nor communicate it to others.

Heuristic 2, involving people for assistance, was focused on those that were socially close rather than who were best placed to correct the issue. For instance, nurse colleagues who were a close friend were preferred, and only after these options failed were others in the hospital care chain or physicians engaged; even if the latter were better equipped to solve the immediate issue.

Antecedents of first-order problem solving

Antecedents to first-order problem solving was primarily due to a lack of time for second-order problem solving and the nurses’ (perceived) low status compared to doctors and managers.

First, “nurses did not have slack time to engage in activities outside their immediate patient care responsibilities” (p129). Nurses regularly had more tasks to perform than time, with many working unpaid overtime to get tasks complete before knocking off.

Further, physicians didn’t always provide necessary info or attention to address the nurses’ concerns. Differences in status between nurses and physicians are well documented, and this study confirmed that finding where at times “doctors … treated nurses as low status workers” (p129).

Examples included how nurses’ insights into patient conditions or treatment plans were ignored, or how nurses were sometimes required “to "prove" that they had indeed encountered a problem and were not just personally incompetent” (p129).

?Consequences of first-order problem-solving

This data showed that nurses “through well-meaning, gratifying efforts of working around problems - ironically contribute to their persistence” (p129).

Firstly, first-order problem solving directly benefitted nurses as individuals; these quick solutions allowed them to continue caring for patients and effectively overcame immediate issues. For instance, taking bedsheets from another ward, which resolved the immediate need but then caused downstream issues for that other ward and didn’t communicate or resolve the underlying issue of a lack of bedsheets.

Importantly, effective problem solving “validated nurses' confidence in their competence and professionalism” (p129).

Second-order problem solving was observed for only 8% of the total resolved issues and even then, most of these were “often opportunistic, weak, and unrecognized as a request for organizational improvement” (p130).

Relaying info about problems was more common than suggesting or implementing solutions to problems. Even here, nurses communicated to the person responsible for the problem/fixing the problem in only 7% of cases.

The authors talked about creating a window of opportunity for second-order problem solving. This window was important because “second-order problem solving had to take place during regular working hours and on the nursing unit” (p130); indicating the importance of creating enough time, motivation, and easy access to key people/resources to engage in second-order problem solving outside of their regular scope of responsibilities.

First-order solutions ranged from resolving annoying things, like the bedsheet example above, to something potentially disastrous like mixing up baby tags.

They note that the “observed predominance of first-order problem solving was that organizational systems got worse or stayed the same” (p132). Notably, “nurses used first-order problem solving in an attempt to navigate through the stream of problems that they encountered during their day but in so doing, contributed to the force of the stream working against them” (p132).

Wrapping up the findings they note:

·????????Second-order problem solving were “almost non-existent” among these front-line workers

·????????Nurses derived a sense of competence, pride and well-being from their ability to solve immediate issues, i.e. first-order problem solving

·????????Not only were nurses proud of their ability to navigate challenges but they were “rewarded for doing so”. That is, the “institutional features of the work environment” reinforced an emphasis on first-order problem solving

·????????Engaging in second-order problem solving required trust and time in others to resolve an issue, and thus eroded a nurses’ confidence in independence and competence

·????????Organisational structures and capacities were rarely available for nurses to effective engage in effective second-order problem solving

·????????“negative cycles of problem-solving behavior - addressing symptoms rather than causes (Senge, 1994) - may be reinforced by the front-line context, in which immediate responsibilities supersede potential future benefits of improvement efforts” (p133)

·????????While fixing these issues were beyond the scope of this study, they note that greater focus on how work is done at the front-line is necessary

·????????Further necessary conditions is that organisations must create available opportunities for second-order problem solving as “an explicit part of their job and enough time allocated for improvement efforts” (p134)

[My thoughts: Although this is healthcare, the findings seem broadly applicable elsewhere like in construction (think about how site supervisors have to frequently engage in first-order problem solving to ensure work continues). Seems logical therefore to find ways to build capacities for second-order improvement into workflows, staffing and the like.

Easier said than done, but this isn’t just setting a KPI for workers to engage in a monthly improvement activity, as this would be another task they need to complete eroding the same finite amount of time, but would rather include removing or substituting existing tasks to create the time; or increasing staffing, or a combination of capacities (workload shedding)]

Link in comments.

Authors: Tucker, A. L., Edmondson, A. C., & Spear, S. (2002). Journal of Organizational Change Management.