Planning – How Does It Work

How Does it Work?

The executive function “planning” is actually comprised of a number of different cognitive processes, including working memory, inhibition, and speed of processing. ​1​

Hayes-Roth & Hayes-Roth Cognitive Model of Planning

In this model, the cognitive “space” where plans are made is conceptualized as a blackboard, which is then broken down into 5 conceptual planes of planning that our minds utilize in order to make and carry out plans. Each plane then has several levels of abstraction.​2,3​

1. Meta-plan Plane:​3​ The planner’s conceptualization of the problem at hand

Expand
  1. Problem Definition: “the planner’s conception of the task. It includes descriptions of the goal, available resources, possible actions, and constraints.”
    1. Problem-solving Model: “how the planner intends to represent the problem symbolically and generate potential solutions.” (i.e., is a certain problem an efficiency or scheduling one?)
    2. Policies: “specify general criteria the planner wishes to impose on his problem solution.”
    3. Solution Evaluation Criteria: “indicates how the planner intends to evaluate prospective plans.”

2. Plan Plane: “actions the planner actually intends to take in the world.”​3​

Expand
  1. Outcomes: “what the planner intends to accomplish by executing the finished plan.”
    1. Designs: “characterize the general approach by which the planner intends to achieve the outcomes.”
    2. Procedures: “specify specific sequences of actions.”
    3. Operations: “specify sequences of more specific actions.”

3. Plan Abstraction Plane: “characterize desired attributes of potential plans.”​3​

Expand
  1. Intention: “stimulates efforts to partition the errands into critical and non-critical sets.”
    1. Scheme: “motivates a search for coherent clusters” of tasks
    2. Strategy: “constrain procedural sequences eventually incorporated into the plan.”
    3. Tactic: i.e. if a planner adopted a tactic “that suggested searching for a short-cut between one errand and the next. This…might lead to the discovery and use of one particular short-cut.”

4. Knowledge-Base Plane: “records observations and computations about relationships in the world which the planner generates while planning.” This plane has two types of planning functions: Situation Assessment: “analysis of the current state of affairs,”and Plan Evaluation: “analysis of the likely consequences of hypothesized actions.”​3​

Expand
  1. Errand: computation of facts about each step of the process (i.e. a single errand during a day of running errands” “to evaluate the plan’s gross feasibility.”
    1. Layout: recognizing spatial (or topical?) clustering of individual errands (tasks) “and, as a consequence, formulate a design organized around clusters.”
    2. Neighbor: if “two planned errands are near one another” then “adopt a procedural decision to sequence those two errands.”
    3. Route: noticing and exploiting short-cuts.

5. Executive Plane: where cognitive resources are allocated for the plan.​3​

Expand
  1. Priority: “decisions indicating preferences for allocating processing activity to certain areas of the planning blackboard.”
    1. Focus: “decisions indicate what kind of decision to make at a specific point in time, given the current priorities.”
    2. Schedule: “indicate which of the currently invoked specialists, satisfying most of the higher-level executive decisions to execute.”

Initial Planning vs. Concurrent Planning

Much research about planning distinguishes between two types. First is initial planning (IP), which is the planning done in advance of a situation, and “are more likely to be goal directed, hierarchical, or total order.” Second is concurrent planning (CP), which happens “on-line” or in the moment.​4​

Initial planning loads heavily on working memory, “but the flexibility of storage in working memory means that different alternatives can be considered at relatively low cost.” Concurrent planning doesn’t rely so heavily on working memory, “however, it also has associated difficulties since it may be hard to backtrack having made a wrong move.”​4​

Several factors likely come into play when determining if one will (or should) utilize initial or concurrent planning.

Expand
  1. Problem complexity. Simple plans can be formulated real-time in the working memory with little problem. However when moderately complex problems are faced, initial planning appears to be more helpful. Interestingly, the same doesn’t seem to hold true for high-difficulty problems.​4​
  2. Problem environment. How people choose to plan for problem solving seems sensitive to various environmental factors, such as the wording of a question, the input from those around them, and the physical environment itself (access to paper or technology that can help a person externalize goals and sub-goals can reduce the stress on working memory and allow for more concurrent planning.​4​
  3. Individual and group differences. There are “strong individual preferences for particular strategies, which may in turn be related to individual differences in such things as spatial reasoning or forward planning abilities.” Whether a situation is new or familiar (or if a person is a novice or expert in a particular domain) will also effect planning strategies. ​4​ Furthermore, “societal, family and peer group norms all influence the cognitive, personality and motivational components of an individual’s planning ability in terms of, for example, an individual’s knowledge base, beliefs, skills and goals.”​5​

Process Model

“The ways in which people work with knowledge, or the mental operations employed in working with this knowledge, are what is meant by the term “processes.” In recent years, a series of studies by Mumford and colleagues (..) have begun to isolate the key cognitive processes contributing to planning performance.”​6​

“More centrally, following the proposition that planning requires the mental simulation of future events, the model assumes the importance of forecasting, with forecasts being refi ned through evaluation of fore- casted outcomes, as indicated in the lower third of the model (…). Commonly, we assume that people’s forecasts are inaccurate (…). However, the evidence provided by Dailey and Mumford ( 2006 ) indicates that when people have requisite experience and intend to act on their forecasts, the accuracy of forecasts improves “​6​

“The model of planning processes presented here holds that planning requires “envisioning” the eff ects of action or mental simulation”​6​

“These forecasts, moreover, are used not only to formulate plans but also to identify monitoring markers for plan execution and to formulate backup plans “​6​

“More specifi cally, the model of planning processes assumes that plans are executed in a conscious, opportunistic, fashion, as per the base of the model at hand (Patalano & Seifert, 1997 ). On the one hand, conscious opportunistic plan execution promotes adaptability through focused processing (Jaudas & Gollwitzer, 2004 ). On the other hand, focused, conscious processing may lead to the loss of plan-irrelevant information”​6​

“This model of planning processing holds, in keeping with earlier observations concerning the basis of planning in case-based knowledge (… ), that the analysis of activated cases to identify causes, resources, restrictions, contingencies, goals, actors, affect, and systems provides the basis for formulation of “local” mental models.”​6​

“Accordingly, a key proposition underlying development of the planning model is that cases are indexed with respect to global mental models. With activation of a global mental model, prototype cases are activated. Activated case prototypes will also be infl uenced by situational appraisals and apparent goals operating in the situation, as well as by information gathered with regard to the situation at hand. Activated case prototypes, typically a relatively small number, will be used to identify critical causes and salient goals (Mumford, Schultz, & Van Doorn, 2001 ). Cross-indexing of cases with respect to causes and goals will give rise to the cases activated for further analysis. Analysis of cases will typically focus on critical case elements, causes, resources, goals, actors, with local mental models being built using this material “​6​

“Moreover, the processes specifi ed in this model are held to operate in a dynamic, interactive fash- ion. Thus, activation of global mental models is held to “feed back” into information gathering, again indicated in Figure 2.1 by the use of bidirectional arrows. Although this model indicates that a complex set of processes underlie planning, planning operations can be readily simplifi ed in this model: if cases are tightly bound to situational cues, only one case is used to build a local mental model, forecasting is limited, and backup plans are not generated (Nutt, 1984 ). Thus, this process model allows for both complexity and simplicity in planning”​6​

“Finally, it should be recognized that different planning tasks, and different task domains, may stress and require effective execution of different processes. For example, scanning may be of great importance for research and development planning (Mumford et al., 2008 ), while forecasting may be of greater importance for leader planning”​6​

“A critical proposition of this model is that forecasting is necessary for planning(“​6​ “The model of planning processes also holds that local mental models provide a basis for forecasts and plan formation. “​6​

“Along somewhat diff erent lines, the model of planning processes under consideration also holds that refl ection on the situation at hand contributes to evaluation of forecasts and planning performance. Some evidence bearing on the importance of refl ection in planning has been pro- vided in a study by Strange and Mumford ( 2005 ). In this study, participants, undergraduates, were asked to formulate plans for leading an experimental secondary school and prepare speeches to be given to students, parents, and teachers describing their plans. These speeches were evaluated for perceived utility and emotional impact. Notably, a manipulation was induced to encourage participants to refl ect on their prior personal experience in secondary school. It was found that refl ection, specifi cally self-refl ection, contributed to the production of plans of greater utility and emotional impact”​6​

“This model of planning processes, however, is not tied to a particular task or a particular domain of planning activity (Baer, 2012 ). This point is of some importance because some processes may prove more important on some tasks, in some perform- ance domains, than others. As a result, some caution is called for in extrapolating this general model of planning processes to planning performance in specifi c tasks domains. Moreover, based on their experience and the demands imposed by the planning task at hand, individuals may or may not execute all of these processes. For example, we do not, typically, forecast drive times to work when planning our daily trips, relying instead on simplifi ed plans based on past experience”​6​

“Along related lines, it should be recognized that the model of planning processes presented in the present chapter was based on a key assumption. More specifi cally, we assumed that plans are formulated using case-based, or experiential, knowledge. Although the available evidence does suggest that plans are typically formulated using case-based knowledge (e.g. Berger & Jordan, 1992 ; Klein, Robertson, & Delton, 2010 ), the possibility does remain that other types of knowledge might be used in plan formation, and, if other types of knowledge are used, then a diff erent model of planning processes than the model presented herein would be indicated”​6​

“To envision the eff ects of future actions, people must forecast. These forecasts, however, are based on mental models. The mental models used to formulate these forecasts, in turn, are based on the analyses of available case-based knowledge. The case-based knowledge analyzed, however, is selected based on general mental models and the inferences drawn from these models with respect to critical causes and critical goals. The critical causes, critical goals, and global mental model employed, however, are based on information gathering, monitoring models, and scanning. Moreover, eff ective execu- tion of these planning processes is held to depend on various “support” processes, such as evalu- ation of forecasts, backup plan formation, and situational refl ection “​6​

Sources:

  1. 1.
    Kliegel M, Martin M, McDaniel MA, Phillips LH. Adult Age Differences in Errand Planning: The Role of Task Familiarity and Cognitive Resources. Experimental Aging Research. March 2007:145-161. doi:10.1080/03610730601177395
  2. 2.
    Hayes-Roth B, Hayes-Roth F. A Cognitive Model of Planning*. Cognitive Science. October 1979:275-310. doi:10.1207/s15516709cog0304_1
  3. 3.
    Hayes-Roth B, Hayes-Roth F, Rosenschein SJ, Cammarata S. Modeling Planning as an Incremental, Opportunistic Process. The Rand Corporation; 1976:26. https://apps.dtic.mil/dtic/tr/fulltext/u2/a071035.pdf.
  4. 4.
    Davies SP. Planning and Problem Solving in Well-Defined Domains. In: Morris R, Ward G, eds. The Cognitive Psychology of Planning. Current Issues in Thinking and Reasoning. Psychology Press; 2004:246.
  5. 5.
    Ward G, Morris R. Introduction to the Psychology of Planning. In: Morris R, Ward G, eds. The Cognitive Psychology of Planning. Current Issues in Thinking and Reasoning. Psychology Press; 2004:246.
  6. 6.
    Mumford MD, Mecca JT, Watts LL. Planning Processes: Relevant Cognitive Operations. In: Mumford MD, Frese M, eds. The Psychology of Planning in Organizations. Routledge; 2015:22.