INT’L. J. AGING AND HUMAN DEVELOPMENT, Vol. 62(2) 175-184, 2006

AGE AND PLANNING TASKS: THE INFLUENCE

OF ECOLOGICAL VALIDITY*

LOUISE H. PHILLIPS

School of Psychology, College of Life Sciences and Medicine,

University of Aberdeen

MATTHIAS KLIEGEL

MIKE MARTIN

University of Zurich

ABSTRACT

Planning ability is important in many everyday tasks, such as cooking and

shopping. Previous studies have investigated aging effects on planning,

looking at either widely used laboratory-based neuropsychological tasks such

as the Tower of London (TOL) or more naturalistic planning tasks, such as

organizing shopping errands. In the current study, we compare the effects of

normal adult aging on both the TOL and a more ecologically valid planning

task, the Plan-a-Day (PAD) task. There was a reliable decline in TOL

planning performance with age, but no significant correlation between age

and PAD planning performance. Age-related variance was partly explained

by variance in information processing speed and education. It is proposed that

in more ecologically valid planning tasks, age changes in processing speed

can be compensated for by task-related knowledge. Implications for everyday

planning performance by older adults are considered.

*Preparation of this article was supported, in part, by a grant from Cusanuswerk, Bonn, Germany,

and a grant from the German Science Foundation DFG (Ma 1895/4-1).

175

� 2006, Baywood Publishing Co., Inc.

Cognitive planning is involved in a range of important life skills, such as cooking,

shopping, and many occupational tasks. There is evidence that executive func-

tions, such as planning, are better predictors of the ability to carry out daily

activities in old age than more traditional cognitive measures, such as intelli-

gence and memory (Cahn-Weiner, Malloy, Boyle, Marran, & Salloway, 2000).

However, although several studies have found age effects on planning tasks

(e.g., Gilhooly, Phillips, Wynn, Logie, & Della Sala, 1999; Kliegel, McDaniel,

& Einstein, 2000), age does not seem to influence all aspects of planning in the

same way (for a review see Phillips, MacLeod, & Kliegel, 2005). One of the most

widely used neuropsychological tasks of planning is the Tower of London (TOL)

task (see Shallice, 1982) in which participants are shown a start and goal set of

disks placed on rods, with the disks differing in position. The task is to make a

mental plan which moves the start set of disks to match the goal in the minimum

moves possible, and then to physically execute that plan. Older adults make less

accurate mental plans than young on the TOL (Gilhooly et al., 1999), as well as

making more excess moves over the minimum necessary to solve TOL trials

(Andrés & van der Linden, 2000).

Lachman and Burack (1983) argue that although age differences are found in

laboratory-based planning tasks, using planning materials that are more familiar

might reduce or eliminate age differences. Recently, evidence has been found to

support this proposal in studies indicating that there are no age differences found in

shopping errand tasks where the material is more realistic (Garden, Phillips

& MacPherson, 2001; Kliegel, Martin, McDaniel, & Phillips, in prep.). Also, a

recent meta-analysis of prospective memory tasks, which require the planning

and execution of an intention to carry out a task (Henry, MacLeod, Phillips &

Crawford, 2004), indicates a substantial age-related deficit in plan execution

on laboratory tasks, but an age-related benefit of the same magnitude in plan

execution in naturalistic tasks. One aim of the current study is to extend these

findings and investigate whether this pattern of age differences in planning on

laboratory tasks along with age stability in more ecologically valid planning

tasks can be found in a single sample of young and old adults. Although there

is often recognition of the importance of looking at the performance of older

adults in more realistic contexts, there is considerable tension between the

development of methods which tap into realistic environments and experimental

control and rigor (Czaja & Sharit, 2003).

The task which we use to investigate more contextualised planning is called

the Plan-a-Day test (PAD) (Funke & Krüger, 1993; Kohler, Poser, & Schönle,

1995) and investigates the ability to plan a work schedule. This task is based

upon classical daily errands tasks (e.g., Hayes-Roth & Hayes-Roth, 1979) and

requires participants to run a number of errands in a fictitious workplace setting,

given specific constraints concerning errand priority and the time course of

the task. This task has previously been used in neuropsychological assessment

(e.g., Gouzoulis-Mayfrank, Thimm, Rezk, Hensen & Daumann, 2002), as well as

176 / PHILLIPS, KLIEGEL AND MARTIN

in occupational settings (e.g., Funke & Krüger, 1993). Funke and Krüger (1995)

also report data on the use of the PAD task in assessment centers and found

acceptable reliability of the task. They conclude that the measure has good

convergent and discriminant validity as a measure of everyday planning in

workplace settings because: a) the task successfully distinguishes managers from

non-managers, b) PAD performance correlates with organizational skills in the

workplace, and c) PAD variance loads on a separate factor compared to other

tests used in assessment centers. This task meets important criteria for ecological

validity in a simulated work task identified by Czaja and Sharit (2003): sampling

of tasks similar to those used in work settings and prediction of actual skills in

the workplace.

A further issue in relation to age differences in planning concerns the involve-

ment of cognitive resources. There is evidence that age-related changes in many

cognitive functions such as reasoning and memory correlate with changes in

relatively simple measures of information processing speed (e.g., Salthouse,

1996). It is generally assumed that age-related changes in executive functions,

such as planning, reflect a specific age change separate from more general

changes in information processing speed. In the current article, we test whether

age variation in planning tasks overlaps with variance in processing speed.

Another potential explanation for age differences in planning is that older adults

have poorer ability to inhibit task-irrelevant material (Hasher, Stolzfus, Zacks,

& Rypma, 1991) and thus, are less able to select and focus upon the most

critical information central to constructing and executing a plan. There is evi-

dence that variance in inhibitory functioning can explain some of the age-related

variance in a complex holiday planning task (Martin & Ewert, 1997). In the

current study, we assess inhibition using the most common measure of this

construct, the Stroop task.

In sum, the first aim of the current study is to investigate the effects of adult

aging on both abstract (TOL) and more contextualised (PAD) planning tasks in

the same population, with the prediction that age differences should be smaller

in the more ecologically-valid PAD task. A further aim is to investigate whether

any age differences found in planning can be explained by variance in processing

speed or inhibition.

METHOD

Participants

Participants were 39 young (M = 24.8 years, SD = 2.0, range = 22–31; 18 male,

21 female) and 39 old (M = 69.5, SD = 5.5, range = 60–80; 10 male, 29 female)

participants who reported no history of vision or hearing difficulties. The younger

adults had more years of education (M = 13.2, SD = 0.9) compared to the older

adults (M = 9.9, SD = 3.1), F(1, 76) = 42.8, p < .01. Thus, education will be

considered in the following analyses.

AGE AND PLANNING TASKS / 177

Procedure

Planning Measures

The Tower of London (TOL) is a traditional laboratory planning task in which

the difficulty can be varied by using different start and end states to manipulate

plan length. In the current computerized version of the task, three different-colored

disks were displayed on three rods that varied in size, and the participants had

to move the disks using the computer keyboard. Participants were instructed

to achieve the end state in the fewest possible moves possible but were not

specifically told to make a full plan in advance. Various task instructions can be

given in the TOL task, but there is evidence that instructions to produce a full

mental plan does not reduce the number of moves needed to solve the TOL task

(Phillips, Wynn, McPherson, & Gilhooly, 2001).The main dependent variable

was the difference between the minimum number of moves (i.e., 50) in which

the 10 given problems could be solved and the number of moves actually made

by a participant. Thus, high scores on the TOL task indicate more excess moves

(i.e., worse planning performance).

The second, more contextualised planning measure was the Plan-a-day test

(PAD). This is a computerized task in which participants have to run a number of

errands in a fictitious setting given specific constraints concerning errand priority

and the time course. Participants are told to imagine themselves as an employee

of a company who has to carry out a number of appointments during a fictitious

day. They are encouraged to carry out as many appointments as possible. The

appointments all take place within the area of the company presented on the

computer screen, which consists of several buildings that are scattered over a

wide area. Participants are informed that each appointment can only be met at

a specific time or in a specific time window. They are also prompted that the

scheduling of these appointments must take into account the distances between

the respective locations. Participants can always view the set of appointments

they have to schedule by pressing a function key. The option to delete plan

elements and modify schedules is also available. Participants are first pre-

sented with a practice trial, and only once they have completed this correctly

do they continue with the “test” part of the PAD which consists of two “days”

for which seven appointments have to be scheduled in 20 minutes test time

each. The dependent variable is the number of errands accomplished weighted

with respect to their priority. Hence, high scores on the PAD task indicate more

accurate planning.

Predictors

Speed of processing was assessed with the Digit-Symbol Substitution Test

(DSST) from the revised Wechsler Adult Intelligence Scale (Wechsler, 1981),

scored as the number of correctly translated digits within a 90s period.

178 / PHILLIPS, KLIEGEL AND MARTIN

Finally, we included a color-word version of the Stroop-task in order to measure

inhibition (e.g., Houx, Jolles, & Vreeling, 1993). In this task, the baseline trials

consisted of four types of color bar (red, blue, green, and yellow) with instructions

to name the colors as fast as possible. The interference trials consisted of the

four color names printed in mismatching colors with instructions again to name

the color each words is written in as fast as possible. Each Stroop condition

began with practice on five items; followed by timed performance on the 20

test items (consisting of five rows of four items each). The dependent variable

was the time difference between the baseline and the interference condition.

Analysis Strategy

Multiple linear regression analysis was used to investigate whether age-related

variance could be explained by measures of education, speed, and inhibition.

A regression equation with age as the only predictor was calculated first. In

order to examine if education, speed, and inhibition could explain age-related

variance, a further hierarchical regression analysis was then performed. In the

hierarchical regression, education was entered as the first predictor, then, DSST

speed, with Stroop inhibition performance entered in a third step and age in a

final step. This allows examination of correlates of the age-related variance and

also addresses whether inhibitory control functions can account for variance in

planning beyond slowed information processing.

RESULTS

Correlations between age and the planning measures are reported in Table 1.

There was a highly significant relationship between age and TOL planning

performance, r = 0.555, p < .001, with age explaining 31% of the TOL variance.

Older adults showed much poorer planning on the TOL task (young: M = 3.5,

SD = 3.4, old: M = 9.6, SD = 5.7; range of scores = 0 to 22). The correlation

between age and PAD planning performance was not significant (young: M = 63.8,

SD = 5.1 versus old: M = 61.3, SD = 5.7; range of scores = 45 to 68), with age

explaining 4% of variance. Correlations between the planning and predictor

measures are also reported in Table 1. These indicate that TOL performance

correlated with education, inhibition, and speed. PAD planning performance did

not relate to inhibition score but did correlate with education and speed.

Next, multiple hierarchical linear regression analyses were conducted to

investigate whether variance in education (Step 1), speed (Step 2), and inhibition

(Step 3) mediated age effects (Step 4) in both planning tasks. Results are sum-

marized in Table 2. For the TOL task, education and speed were both signifi-

cant predictors of performance, with no remaining variance being explained

by inhibition. However, even after considering education, speed, and inhibition,

a small part of the age-related variance remained unexplained. Comparing the

AGE AND PLANNING TASKS / 179

180 / PHILLIPS, KLIEGEL AND MARTIN

Table 2. Regression Analyses Predicting Performance

on Planning Tasks

Planning task

TOL PAD

�R2 p-Value �R2 p-Value

Step 1: Education

Step 2: Speed

Step 3: Inhibition

Step 4: Age

.124

.133

.000

.054

.002

.001

.889

.022

.081

.023

.000

.016

.013

.173

.917

.263

Overall shared variance,

Age and Planning .308 .000 .041 .081

Note: TOL = Tower of London planning task, PAD = Plan A Day task, Stroop = Difference

score between Stroop inhibition and control tasks, DSST = Digit Symbol Substitution Task.

Table 1. Correlations between Planning Measures and

Predictor Variables

Age Education TOL PAD Stroop

Education

TOL

PAD

Stroop

DSST

–.600**

.555**

–.202

.553**

–.860**

–.352**

.285*

–.412**

.591**

–.226

.309**

–.502**

–.171

.291* –.593**

Note: TOL = Tower of London planning task, PAD = Plan A Day task, Stroop = Difference

score between Stroop inhibition and control tasks, DSST = Digit Symbol Substitution Task.

*p < .05. **p < .01.

remaining age related variance in TOL planning (see Table 2, Step 4) with the

initial age variance (see Table 2, final row) reveals that education and speed

together explain around 83% of the age-related variance in TOL. For the PAD

task, education was the only significant predictor, with speed no longer explaining

variance in the hierarchical analysis.

DISCUSSION

The results from this study support the prediction that age differences in a

laboratory-based planning task will be much greater than on a more contextual-

ized task. Within the same group of participants there were substantial age

differences in planning efficacy in the TOL task, but no significant effect of age

on the PAD task, despite the fact that the PAD task was rather complex and

required simultaneous consideration of multiple task elements and constraints.

Both tasks showed significant correlations with a processing speed measure,

and most (but not all) of the age variance in the TOL was explained by con-

trolling for speed and education. These results suggest that both the TOL and

PAD tasks are influenced by speed of processing, which decreases substan-

tially with age. However, it seems likely that contextual background may be

more important in the PAD task. It is plausible that some form of compen-

satory mechanism may be in operation whereby older adults are able to make

use of their previous knowledge about real life scheduling to override the influ-

ence of slowed processing speed on the PAD (see e.g., Marsiske, Lang, Baltes,

& Baltes, 1995). This is supported by other research (Kliegel et al., in prep)

indicating that in tasks involving familiar materials older adults are good at

selectively attending to task-relevant information, and this can compensate for

resource changes with age.

It is worth noting that even the more ecologically-valid PAD task used here

demanded the use of computer skills which are likely to be underdeveloped in an

older adult sample. The task therefore may still overestimate any age-related

declines in planning, due to the more unfamiliar format for older adults. Note

though that computerized testing is used for many studies of age effects on

ecologically valid tasks; Czaja and Sharit (2003) argue that because most current

workplace settings involve computer interaction, it is appropriate for simulated

work sample tests to be computerized. However, it would be useful in future

research into ecologically valid planning tasks in aging to: a) assess computer

abilities and familiarity, b) include a non-computerized measure of planning

if possible, or c) recruit young and old participants with equivalent levels of

computer experience.

The finding that age differences in TOL performance can be explained by

considering processing speed but not inhibition scores indicates that the

effects of age on this task may reflect a more global cognitive change rather than

a specific executive function deficit. This fits with findings that the age changes

AGE AND PLANNING TASKS / 181

in TOL planning performance are not differentially larger than changes in

intelligence test performance (Crawford, Bryan, Luszcz, Obonsawin, & Stewart,

2000) and dual task results, which indicate that in younger adults the TOL

specifically loads executive function while in older adults the TOL loads

more general cognitive resources (Phillips, Gilhooly, Logie, Della Sala, &

Wynn, 2003).

In conclusion, these results have implications for the interpretation of age

differences in planning tasks. It cannot be assumed that age differences on

laboratory tasks, such as the TOL, imply poor ability to plan in the real world,

because knowledge-based compensatory mechanisms may be in place which

facilitate performance on more realistic tasks. Nor can it be assumed that age

differences on the TOL reflect a specific executive deficit of cognition—instead

it may be more parsimonious to interpret the majority of evidence on age effects

of the TOL as indicative of a more global cognitive change with age. Finally,

we propose that on planning tasks which deal with contextualised materials

older adults can compensate for age-related declines in processing speed through

utilization of relevant knowledge.

ACKNOWLEDGMENT

We would like to thank Professor Margie Lachman for her comments on the

results reported here.

REFERENCES

Andres, A. U., & van der Linden, M. (2000). Age-related differences in super-

visory attentional functions. Journal of Gerontology: Psychological Sciences, 55,

373-380.

Cahn-Weiner, D.A., Malloy, P.F., Boyle, P.A., Marran, M., & Salloway, S. (2000).

Prediction of functional status from neuropsychological tests in community-dwelling

elderly individuals. Clinical Neuropsychologist, 14, 187-195.

Crawford, J. R., Bryan, J., Luszcz, M. A., Obonsawin, M. C., & Stewart, L. (2000).

The executive decline hypothesis of cognitive aging: Do executive deficits qualify

as differential deficits and do they mediate age-related memory decline? Aging

Neuropsychology and Cognition, 7, 9-31.

Czaja, S. J., & Sharit, J. (2003). Practically relevant research: Capturing real world

tasks, environments and outcomes. The Gerontologist, 43, Special Issue 1, 9-18.

Funke, J., & Krüger, T. (1993). “Plan-A-Day” (PAD). Bonn: Psychologisches Institut

der Universität Bonn.

Funke, J., & Krüger, T. (1995). “Plan-A-Day”: Konzeption eines modifizierbaren Instru-

ments zur Führungskräfte-Auswahl sowie erste empirische Befunde [Concept and first

empirical results of a modified management assessment test]. In J. Funke & A. Fritz

(Eds.), Neue Konzepte und Instrumente zur Planungsdiagnostik [New concepts and

instruments for the assessment of planning abilities] (pp. 97-120). Bonn: Deutscher

Psychologen Verlag.

182 / PHILLIPS, KLIEGEL AND MARTIN

Garden, S., Phillips, L. H., & MacPherson, S. E. (2001). Mid-life aging, open-ended

planning and laboratory measures of executive function. Neuropsychology, 15,

472-482.

Gilhooly, K. J., Phillips, L. H., Wynn, V. E., Logie, R. H., & Della Sala, S. (1999). Planning

processes and age in the 5 disc Tower of London task. Thinking and Reasoning, 5,

339-361.

Gouzoulis-Mayfrank, E., Thimm, B., Rezk, M., Hensen, G., & Daumann, J. (2002).

Memory impairment suggests hippocampal dysfunction in absintent ecstasy users.

Progress in Neuropsychopharmacology and Biological Psychiatry, 27, 819-827.

Hasher L., Stolzfus, E. R., Zacks, R. T., & Rypma, B. (1991). Age and inhibition. Journal

of Experimental Psychology: Learning Memory and Cognition, 17, 163-169.

Hayes-Roth, B., & Hayes-Roth, F. (1979). A cognitive model of planning. Cognitive

Science, 3, 275-310.

Henry, J. D., MacLeod, M., Phillips, L. H., & Crawford, J. R. (2004). A meta-analytic

review of age effects on prospective memory. Psychology and Aging, 19, 27-39.

Houx, P. J., Jolles, J., & Vreeling, F. W. (1993). Stroop interference: Aging effects assessed

with the Stroop color-word test. Experimental Aging Research, 19, 209-224.

Kliegel, M., Martin, M., McDaniel, M. A., & Phillips, L. H. (in prep.). Older adults’

planning: Experience can compensate for cognitive decline. Manuscript in preparation.

Kliegel, M., McDaniel, M. A., & Einstein, G. O. (2000). Plan formation, retention, and

execution in prospective memory: A new paradigm and age-related effects. Memory

and Cognition, 28, 1041-1049.

Kohler, J. A., Poser, U., & Schönle, P. W. (1995). The application of “plan-a-day” in

neuropsychological assessment and therapy. In J. Funke & A. Fritz (Eds.), Neue

Konzepte und Instrumente zur Planungsdiagnostik (pp. 167-181). Bonn: Deutscher

Psychologen Verlag.

Lachman, M. E., & Burack, O. R. (1983). Planning and control processes across

the lifespan: An overview. International Journal of Behavioral Development, 16,

131-143.

Marsiske, M., Lang, F. R., Baltes, P. B., & Baltes, M. M. (1995). Selective optim-

ization with compensation: Life-span perspectives on successful human develop-

ment. In R. A. Dixon & L. Bäckman (Eds.), Compensating for psychological deficits

and declines: Managing losses and promoting gains (pp. 35-79). Mahwah, NJ:

Erlbaum.

Martin, M., & Ewert, O. (1997). Attention and planning in older adults. International

Journal of Behavioral Development, 20, 577-594.

Phillips, L. H., Gilhooly, K. J., Logie, R. H., Della Sala, S., & Wynn, V. (2003). Age,

working memory, and the Tower of London task. European Journal of Cognitive

Psychology, 15, 291-312.

Phillips, L. H., MacLeod, M., & Kliegel, M. (2005). Adult aging and cognitive planning.

In G. Ward & R. Morris (Eds.), The cognitive psychology of planning (pp. 111-139).

Hove, UK: Psychology Press.

Phillips, L. H., Wynn, V. E., McPherson, S., & Gilhooly, K. J. (2001). Mental planning

and the Tower of London task. Quarterly Journal of Experimental Psychology A, 54,

579-598.

Salthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition.

Psychological Review, 103, 403-428.

AGE AND PLANNING TASKS / 183

Shallice, T. (1982). Specific impairments of planning. Philosophical Transactions of the

Royal Society of London B, 298, 199-209.

Wechsler, D. (1981). Wechsler Adult Intelligence Scale–Revised (WAIS-R). New York:

Psychological Corporation.

Direct reprint requests to:

Louise H. Phillips

School of Psychology

College of Life Sciences and Medicine

University of Aberdeen

Aberdeen

AB24 2UB

Scotland, UK

e-mail: louise.phillips@abdn.ac.uk

184 / PHILLIPS, KLIEGEL AND MARTIN