A Computerized TOL

Published on August 3, 2024

Lifespan Performance on a Computerized Version of the Tower of London DX: From 5 to 89 Years of Age

Ana Levy, Devin Levy & Hasker Davis

CogQuiz LLC, Los Angeles, California

Abstract

The Tower of London-Drexel 2nd Edition (TOLDX) is a commercially available tower transfer task designed to assess planning ability by manipulating spheres on a wooden peg board. In order to solve a problem on the TOL efficiently one must successfully create and implement a successful problem solution. The proposed study will assess the utility of a computerized version of the TOLDX on a population ranging in age from 16-89 years of age. Multiple regressions and correlations will be run on a sample of 759 participants to assess how years of education, sex, age, and fluid intelligence affect the total excess moves, total initial time, and total completion time across ten trials on the TOLDX. Additional univariate analyses will be run on the outcome measures to detect the effect of age on the outcome measure gathered during testing.

Introduction

The Tower of London (TOL) is a commonly used tower transfer task originally designed to detect and measure planning deficits in individuals with frontal lobe lesions. In the manual TOL, participants are shown two identical stimulus boards containing three colored spheres: red, white, and green mounted on three wooden pegs of varying size able to hold up to one, two, or three spheres. At the beginning of each trial, the participant’s board is reset to the same start arrangement while the experimenter’s board displays the goal arrangement. The participant’s goal is to match their arrangement (start state) to the arrangement displayed on the experimenter’s board (goal state). Measures of performance on the TOL vary between studies but the most commonly used metrics of success are trial completion (creating the correct arrangement), total number of moves, and total completion time (Berg & Byrd, 2002; Berg, Byrd, McNamara, & Macdonald, 2006). Additional measures of performance will be discussed later in the paper.

The TOL was adapted from another tower transfer task: the Tower of Hanoi (TOH) (Shallice, 1982). Both of these tests are similar in the sense that the participant must use their planning abilities to find and execute the most efficient means of moving a series of spheres or discs from a starting state to a goal state that has been specified by the examiner (Berg & Byrd, 2002). However, in recent years the TOL has become a much more commonly used test than the TOH due to its wider and more versatile problem set and the greater number of testing parameters that are measured in the TOL.

Methods

The TOL has a diverse problem set that exposes participants to a wide variety of different tower arrangements. The TOH has the same arrangement across various trials so that as soon as the “trick” to solving the problem is discovered, individuals simply need to refine this strategy in subsequent trials to eventually hone their performance. There is no universal iterative strategy underlying successful solution of TOL problems. Thus, the individual must utilize a greater assortment of different strategies for each trial. There may be some overlap in the strategies used to solve TOL problems. Some problem sets utilize “problem nesting” which is the process of embedding moves from a 3-move problem into a 5-move problem later in the test (Berg & Byrd, 2002). Nesting a problem allows the researchers to add additional moves that will place a greater strain on the individual’s working memory while retaining similar planning issues or strategies. Additionally, different TOL arrangements allow the experimenter greater control over how difficult the problem set will be and to systematically increase problem difficulty. As well as controlling for difficulty, the experimenter can create a wider variety of qualitatively different problems (Berg & Byrd, 2002; Shallice & Burgess, 1991). Having a wider variety of potential problems allows greater applicability to a variety of populations with varying levels of impairment. Since there are numerous arrangements of spheres and pegs available, experimenters are able to control the difficulty of their test as the problem sequence progresses. This allows experimenters to tailor the test specifically to the population of interest (Kaller et al., 2012; Unterrainer & Owen, 2006). For example, Köstering et al. (2015) found that participants who were in the cognitively impaired sample scored consistently lower than the participants in the healthy sample.

Results

When Shallice originally created the TOL, he proposed that inhibition would be one of the most important factors contributing to performance. When working on a problem, individuals are likely to use what he referred to as Contention Scheduling (Shallice, 1982). He proposed that contention scheduling is a fast and automatic process by which an individual will quickly view and process an unfamiliar set of visual stimuli and the mind will rapidly select what it believes to be the most relevant schema to fit the current situation. In relation to the TOL, a tower arrangement will be viewed and the mind will select the schema or set of actions that it believes to be most efficient for that problem. The strongest and most readily accessible schema is also the most familiar. The reason Shallice thought inhibition was so important was because as participants progress through the test, they are exposed to more and more problems and begin to formulate numerous solution schemata. Though the trials may be similar in appearance, the optimal solution may be different, and the participants must inhibit non-optimal schemata from becoming activated which could lead to an incorrect or inefficient problem solution. Shallice cited this as one of the main reasons that individuals with frontal lobe lesions performed poorly on the TOL. They were not able to take into account all of the visual stimuli in each problem and as a result, inefficient schemata were activated and followed when solving various problems.

Discussion

There are many different forms of the TOL; however, the present study will focus on a computerized version of the Tower of London-Drexel 2nd Edition (TOLDX). This test was originally adapted from the Shallice TOL by Culbertson & Zillmer (1998) as a test of executive function in children and was later modified to assess executive functioning in adults with ADHD (Riccio, Wolfe, Romine, Davis, & Sullivan, 2004). The manual TOLDX is a commercially available test consisting of 10 trials ranging in moves from 4 to 7. While many of the rules remain the same, several changes were made to improve upon the original TOL. The first of these changes was to remove repeated trials which were used to detect how many times a participant would need to try a problem before they solved it using the minimum number of moves or as it will be referred to here, the optimal path. By eliminating the repeated trials, the test maintained a higher level of novelty, reduced practice effects, and provided a shortened testing time (Riccio et al., 2004).

The next step was to include 6 and 7 move problems to reduce the ceiling effect and increase test sensitivity. In their studies with the TOL, Kaller, Unterrainer, and Stahl (2012) reported that the use of 7 move problems helps to identify the “upper range” of planning ability. This concept is further supported by Albert and Steinberg’s (2011) findings that the greatest difference in planning ability was found on the most difficult problems with the greatest number of moves. The use of higher move problems allows the users of the TOLDX to more accurately differentiate between high and low planning performance.

Arguably the most significant improvement in the TOLDX was the standardization of the test and the development of a comprehensive normative base. A detailed set of instructions for administration, scoring, and interpretation were created to reduce the variability among administrations of the TOLDX and to increase the comparability of studies (Culbertson, Moberg, Duda, Stern, & Weintraub, 2004; Culbertson & Zillmer, 2001). A common thread in the TOL literature is that a variety of different forms of the TOL have been used (“balls and pockets” and different problem sets) and performance is analyzed using different outcome measures (initial time and excess moves vs total trial completions and rule violations). As the inter-test variability increases, it becomes progressively more difficult to compare findings, since different parameters tap into different cognitive processes. By standardizing the procedure, researchers are able to investigate the same cognitive processes with different populations, showing how various conditions can differentially affect planning ability.

After creating the test and the administration guide, Culbertson and Zillmer (2001) established expansive, lifespan norms. The TOLDX was normed using a sample of healthy American and Canadian individuals ranging in age from 7 to 80 (n = 990). These were then broken into 9 different age brackets; 7-8 (n = 110), 9-10 (n = 157), 11-12 (n = 103), 13-15 (n = 76), 16-19 (n = 162), 20-29 (n = 192), 30-39 (n = 74), 40-59 (n = 77), and 60-80 (n = 39). Norms were also established for children with ADHD using two different clinical child populations ranging in age from 7-15 (n = 129; n = 115). They reported norms on a set of outcome measures used in the TOLDX, further increasing comparability across studies. These measures include total move score (total number of moves to solve a problem), total correct score (number of problems solved using optimal path), total time violations (number of times participant exceeds 1 and 2 minute limits), total rule violations (total move errors committed on a problem), total initiation time (time between first exposure to problem and first move), total execution time (time spent physically moving spheres after the initial think time), and total problem solving time (total completion time across 10 trials). These measures will be discussed in greater detail below.

While the TOLDX was developed as a manual test to be implemented using wooden stimulus boards, there is an increasing amount of evidence to support the use of computerized methods of testing. Using computerized testing methods allows for greater standardization of testing procedures and can reduce experimenter biases that may come out during an in person testing session. Additionally, since the test has been standardized, experimenters do not need to be as highly trained as would be necessary in a traditional testing format (Zygouris & Tsolaki, 2015). Since the experimenters do not need to be trained as thoroughly, overall cost and time of the experiment can be reduced. Research assistants do not need to be trained compensated. As well as ease of administration, computerized testing simplifies the data gathering process and, in many cases, can increase the accuracy of the measurement. Since the administration and data gathering and analysis all take place on the same platform, the data can be stored between tests and changes in performance can be easily tracked over time (Cambridge Cognition, 2012; CNS Vital Signs, 2012; Neurotrax Corporation, 2003;).

While there are many benefits to using computerized testing, there are certain risks that come along with the use of technology. One of the biggest concerns is their utility and validity in the screening of older adults. Older adult populations may not be as familiar with some of the interfaces used in these studies (Zygouris & Tsolaki, 2015) and their computer illiteracy can cause anxiety and create confounds in the data (Tierney & Lermer, 2012). While these are all valid risks that must be taken into account when designing a study, there is evidence that this population is becoming more computer literate. Hart, Chapparo, and Halcomb (2008) report that older adults are the fastest growing demographic of internet users and younger adults are aging in an increasingly technological world. The older adults of the future will be increasingly familiar with the type of technology used in testing and taking the steps now to establish psychometrically sound measures is increasingly important.

Since its creation as a means of assessing planning ability, the TOL has been implemented as a tool for measuring various domains of executive functioning (Bottari et al., 2009; Köstering et al., 2015; Owen, 2005) as well as spatial planning (Berg & Byrd, 2005; Berg, Byrd, McNamara, & MacDonald, 2006; Kaller, Unterrainer, & Stahl, 2012; Pulos & Denzine, 2005; Shallice, 1982; Unterrainer & Owen, 2006), working memory (Albert & Steinberg, 2011; Berg & Byrd, 2002; Pulos & Denzine, 2005), inhibition (Albert & Steinberg, 2011; Berg & Byrd, 2002; Shallice, 1982), and task shifting (Pulos & Denzine, 2005).

In addition to Shallice’s (1982) original findings of deficits in individuals with frontal lobe lesions, there have emerged several additional populations that frequently show deficits in their performance on the TOL; those with neurological and psychiatric conditions such as depression or Parkinson’s (Jacobs & Anderson, 2002), individuals who have suffered a traumatic brain injury (TBI), Alzheimer’s and related dementias (Carlin et al., 2000), ADHD (Culbertson & Zillmer, 1998), Autism (Wisley & Howlin, 2009), and schizophrenia (Landua & Morris, 2011).

In previous studies the TOLDX has shown an ability to differentiate between healthy and clinical populations. Culbertson et al. (2004) conducted a study using 65 patients in an outpatient setting who had been seeking assistance for Parkinson’s disease (PD) and 34 healthy, demographically matched controls. They found the TOLDX was an effective means of detecting executive function deficits in individuals with PD who required a greater numbers of moves and more time to solve each problem than controls. This finding remained even after they controlled for motor deficits. A similar finding was recorded by Krishnan, Smith, and Donders (2012) in their study of TOLDX¬ performance with adults who suffered a TBI. Individuals with a TBI, regardless of severity, scored reliably worse than the healthy control group.The deficits were mostly in the number of movesand execution time, as opposed to overall time, and were posited to be due to the use of inefficient problem solving strategies.

TOL studies have employed a variety of measures to quantify planning. Berg et al. (2006) highlighted the most commonly used measures of TOL performance and categorized them into three main groups; success/accuracy of the solution, efficiency of the problem, and speed of performance and planning when achieving a solution.

The success/accuracy of solution records how many problems have been successfully solved across the entirety of the test. This parameter is limited by either a certain number of moves or a set time limit. All solutions that are completed within the given parameters are considered perfect solutions and incomplete problems are defined as imperfect solutions. At the end of the test, the percent of perfect solutions are recorded. These studies can be informative as there is a significant loading on the individual’s planning abilities since they must efficiently plan out their solution to complete the problem in the given time or move limit.

The efficiency of solution, is closely related to the previous measures in that problems are either time-limited or performance-limited. In time-limited problems, participants are allowed a certain amount of time to complete each problem and the number of moves needed to complete the problem are recorded. These types of tests are most commonly used in the literature and allow the experimenters to look into the total number of excess moves, which is the number of moves, over and above the minimum moves necessary for a perfect solution. The number of excess moves measure allows researchers to investigate the efficiency of problem solution across problems with varying numbers of minimum moves (Berg & Byrd, 2002). Similar to the efficiency parameter, performance-limited problems record how many times it took a participant to solve the trial with a perfect solution. Additional attempts can be seen as excess attempts and may indicate perseverative planning errors.

The final and most commonly reported TOL measure is the speed of performance and planning during solution. These measures pertain to the speed of various stages of TOL solution, predominantly planning and execution.

The first temporal measure, which has received an increasing amount of attention, is the initial/think time, which it the time spent prior to making the first move. This measure is considered to be the time spent thinking about and planning the problem solution prior to making the first move. The initial time has been shown to increase as problems become harder and the minimum number of moves increases. It is proposed that in cases in which individuals have a longer think time, but subsequently poor performance, the initial time can be indicative of poor planning ability and that these longer initial times were due to problems with formulating an efficient solution (Berg & Byrd, 2002; Berg et al., 2006). Albert and Steinberg (2011) and Ward and Allport (1997) found that initial time was negatively correlated with total moves such that those who spent more time planning prior to beginning the problem solved the problem in fewer moves.

Given the importance that Shallice placed on inhibition as an indicator of success, Culbertson and Zillmer (2001), Steinberg et al. (2009), and Albert and Steinberg (2011), proposed that initial time could be a measure of inhibition. While they still believed that this time was spent creating and planning out their solution, they added that during this planning phase, participants must also inhibit any sub-optimal solutions, or schemata, that may have been activated by the presented problem arrangement. In addition to being the time spent thinking, it could also be a measure of the time spent inhibiting sub-optimal solutions.

The final two temporal measures discussed by Berg et al. (2006) are total time and execution time. Total time can be a useful measure, especially in performance-limited problems, because it can reveal difficulties in the execution of mental plans that have been made. However, it can also be difficult to interpret the full meaning of this measure since it is comprised of all aspects of the problem solution; the time spent planning, making moves, and pausing to update the problem’s solution. One way researchers have sought to remedy this issue is by looking at the execution time, the time spent working on and solving the problem. This number can be easily attained by subtracting the initial time from the total solution time.

As the literature on the TOL has grown, researchers have begun to discover several factors that have a significant effect on performance. These factors include age, arrangement of the boards during a test, and execution strategies that are used to solve to the problems.

Age has repeatedly been shown to be a moderator of performance on all measures of planning ability on the TOL (Albert & Steinberg, 2011; Berg, Byrd, McNamara, & Case, 2010; Köstering, Stahl, Leonhart, Weiller, & Kaller, 2014; MacLeod & Kliegel, 2005). Various studies have found that on average, older adults (age > 60) had a greater number of total moves and excess moves per trial and they spent a greater amount of time planning than younger participants (Andres & van der Linden, 2000; Bugg, Zook, DeLosh, Davalos, & Davis, 2006). Additionally, Köstering et al. (2014) found an age-related decline in planning performance, as demonstrated by a decrease in the total number of problems solved. They found that this age effect was significant even when the sample size was decreased and compared to other older adults within a 29 year age range (ages 60-89).

Berg et al. (2010) found that both familiarity with the test and minimum number of moves required for a successful solution were both strongly related to performance on the TOL. As the minimum number of required moves increased, performance worsened and completion timesand number of excess moves increased. In contrast, when familiarity increased, performance improved with participants needing fewer moves and completing a greater number of problems with perfect solutions. Also, as individualsgained greater TOL experience the minimum number of moves had a diminished effect on overall performance. They hypothesized that this could be due to two separate factors. With increased familiarity, participants becomes less concerned with finding and planning the most efficient solution and instead focus onlearning the cues that can be extracted from the arrangement of the problem. In reference to Shallice’s (1982) idea of contention scheduling, participants are creating more and more schema that can be rapidly activated by certain visual cues within the problem, thus reducing their need to inhibit inefficient strategies. Berg et al. (2010 also postulated that as participants gain more experience, the importance of solving the problem with the minimum number of moves may become less important. This is an interesting phenomenon as instructions commonly include a line about solving the problem in as few moves as possible or trying to find the optimal solution. An investigation into attitudes toward testing is beyond the scope of the proposed test.

The final parameter relates to execution strategies that are utilized throughout the solution. The most commonly studied strategy is sub-goaling or “chunking.” Sub-goaling is the process of breaking each problem into smaller more manageable pieces and solving each of these sub-goals one at a time as the test taker progresses through the problem. Sub-goaling employs online planning which involves constantly updating the mental image of the problem’s solution as one progresses through their solution. In line with the theories proposed about initial time, Kaller, Unterrainer, Rahm, & Halsaband (2004) found that initial times were longer for those who used sub-goaling during their solution.

Several cognitive processes that are considered important to TOL planning and goal completion are the following: 1) the ability to recognize what the goal state looks like and acknowledging when it has been achieved, 2)anticipating the outcome of various behaviors and actions in relation to goal-attainment (Carlin et al., 2000; Kaller et al., 2004), 3) generating and storing mental representations of the process of events that leads from the initial action to goal completion (Unterrainer & Owen, 2006), and 4) the inhibition of distracting and irrelevant stimuli and preventing impulsive, unplanned responses (Shallice, 1982; Unterrainer & Owen, 2006). Planning requires one to create mental images and representations of the problem, the moves necessary to complete the goal, and being able to mentally extrapolate how these moves will affect the overall goal process. As a result of this, one of the most important aspects of planning is a highly functioning working memory.

Working memory (WM) is responsible for creating the mental representation, or problem space, where the various components for TOL goal completion are generated and, more importantly, organized into a cohesive order (Pulos & Denzine, 2005). Creating a problem space assists the participant planning out the optimal path and visualizing the effect of incorrect moves that can be made throughout the problem (Berg & Byrd, 2002). In keeping with the importance of inhibition, Baddeley (1996) stated that a common aspect of WM is the capacity to focus attention on a specific task, despite the presence of irrelevant stimuli and interference. Performance on the TOL loads heavily on WM and participants must utilize their WM to create a problem solution and to assess how each move could positively or negatively affect goal attainment. These solutions are generally created at the beginning of the problem, accounted for by the initial time, and updated as solution progresses. This becomes increasingly important as the minimum number of moves, and the need for sub-goaling increases, usually starting in 3 move problems and above (Cheetham, Rahm, Kaller, & Unterrainer, 2012). These sub-goals or “chunks” allow for the creation of much shorter mental solutions where the participant is only concerned with two or three moves at a time, instead of attempting to create a solution that will solve the entire problem at once. Hayes-Roth and Hayes-Roth (1979) refer to this simultaneous process of creating and executing a solution as online planning. Through the use of on online planning and chunking, the number of moves held in the problem space at a given time is decreased and the load on the WM is reduced. Using these strategies and decreasing the strain on the WM grows increasingly important as the participant increases in age.

Several studies of WM have shown that like planning, WM develops over time, plateaus, and eventually decreases in old age (Hedden & Gabrieli, 2004; Raz & Kennedy, 2009). Alloway and Alloway (2013) conducted a cross-sectional lifespan study with participants ages 5-80, using an online WM test (Automated Working Memory Assessment). They observed considerable change in WM between the ages of 5 and 19, followed by much slower growth into the twenties, with peak performance in the thirties, and a gradual decline from the 50s until 80 years of age. It was also noted that individuals in their 60s performed similarly to those in their 20s. These results would indicate that there is an inverted U in WM performance across the lifespan with the greatest period of learning taking place in early to mid-adolescence, declining in older age, and peaking during early adulthood. Numerous studies have found similar patterns of WM development and deterioration with the greatest changes in performance happening in early to mid-adolescence and the most significant deterioration beginning in old age (ages 50-60) (Albert & Steinberg, 2011; Cowan, Naveh-Benjamin, Kilb, & Saults, 2006; Fandakova, Sander, Werkle-Bergner, & Shing, 2014).

Aging has also been theorized to affect TOL performance due to a decrease in reaction times which may in turn be due to a decrease in motor control that also comes with healthy aging. However, Salthouse (1985) hypothesized that this could be due to a decrease in the speed and accuracy with which the individual is able to process the environmental/problem-based stimuli, known as visuospatial processing. Visuospatial processing is one’s ability to observe objects within an environment and assess their spatial relationships to other objects (Joyce & Robbins, 1991; Morice & Delahunty, 1996). Towse, Hitch, and Hutton (2002) added to this theory with their findings that speed of processing was generally associated with changes in WM, both in regards to its early adolescent development and its decline in older age. Spatial processing has been closely tied with TOL performance and general planning ability with completion time being highly correlated with visuospatial WM (Gilhooly et al., 2002; Pulos & Denzine, 2005).

In an exploratory factor analysis Gilhooly, Wynn, Phillips, and Della Salla (2002) showed that TOL performance had a high factor loading on visuospatial WM. The importance of visual and spatial abilities on TOL performance was further shown by Pulos and Denzine (2005). Individuals with poor visuospatial processing and WM could not effectively differentiate the novel aspects of the current problem from previously viewed trials. This will lead to the activation of previous schemata and subsequently non-optimal problem solutions.

All of these processes; visual processing, spatial orientation, and speed of processing have been grouped into one overarching construct called fluid cognitive skills. As previously discussed, all of these processes are susceptible to normal, healthy, aging with a slow but steady decline commonly starting around the age of fifty and a much more progressive decline occurring after age sixty-five (Finkel, Reinold, McArdle, & Pederson, 2003; McArdle, Ferrer-Caja, Hamagami, & Woodcock, 2002). The decline in these skills has been posited to be the result of inefficient inhibition of stimuli that is incorrect and can mislead the individual and cause them to pursue ineffective solutions (Albert & Steinberg, 2011).

Rationale  Construct and criterion-related validity has been established for the TOLDX through comparisons with various other problem solving, neuropsychological, and cognitive tests (Culbertson and Zillmer, 1998a; 2001; Larochette, Benn, & Harrison, 2009). Riccio et al. (2004) found significant correlations between the TOLDX and processing speed, perceptual skills, matrix reasoning, and immediate memory. In regards to executive functions, the TOLDX was found to be inversely related with general intelligence scores such as the Wechsler Intelligence Scale for Children (WISC) and various sections of the Wechsler Adult Intelligence Scale (WAIS) (Culbertson & Zillmer, 2001; Culbertson et al., 2004; Sari & Culbertson, 2001). These results would indicate that the TOLDX is a measure of the specific executive function of planning, as opposed to general intelligence. Similarly the TOLDX times have been significantly correlated with the WAIS-III processing speed index, the Wechsler Memory Scale test of working memory, as well as switching tasks such as the completion time on trial B of the Trail Making Test and a Stroop Task (Albert & Steinberg, 2011; Riccio et al., 2004). The TOL can also be considered a switching task because it is necessary to switch between mentally planning the moves and physically moving the pieces into the correct position.

In addition to strong construct validity, the TOLDX has also been shown to demonstrate acceptable temporal stability (Culbertson & Zillmer, 1998b). Culbertson and Zillmer (2001) tested the test-retest reliability of the TOLDX in two clinical populations; adults with PD and children with ADHD. With an interval of approximately 140 days, the PD population displayed reliability coefficients in the high range (r = .59 to r = .81). These scores were highest for total moves and time-related scores. In the child ADHD population, scores were all acceptable, ranging from r = .67 to r = .80, with total moves having the highest test-retest reliability.

The cross-sectional, lifespan study that is being proposed would further demonstrate the age effects that are present in the measures of performance on the TOL and the constructs they are assessing. Since the creation of the TOLDX, only one additional study has created as comprehensive a normative base as that of Culberston and Zillmer (2001). The present study would seek to fill this gap in the literature and provide a newer base of normative data on the TOLDX. Albert and Steinberg (2011) investigated the development of planning ability in a sample of individuals ages 10 to 30 (n = 890). They discovered that improvements in the easiest problems stopped at around age 17 and improvements in the harder problems continued into their 20s with performance topping out around 22-25.

Executive functions and planning are essential components to efficient TOL solutions. As such, the Frontal Hypothesis of Cognitive Aging suggests that the aging process predominantly affects areas that are mediated by the frontal lobes such as general executive functioning and control of behavior (Moscovitch & Winocur, 1995; West, 1996). Given that planning is an important component of executive functioning, the TOLDX study proposed can be a means of further supporting this hypothesis as TOL performance has been shown to be greatly affected by age (Phillips, MacLeod, & Kliegel, 2005). The present study expects to see TOL performance have an age effect with performance being lowest in adolescents and ok, lder adults and peak performance occurring in the early adulthood.

Similar to the original lifespan study, the present study will break the participants into 9 different age groups. However, the current sample will have smaller age ranges which will allow for a more accurate representation of the change in planning ability in older adults. Whereas Culbertson and Zillmer divided their sample into 40-59 years and 60-80, this study would break the sample into the 40s, 50s, 60s, 70s, and 80s. In addition to these smaller groups, the age range will be expanded as well. Köstering et al. (2014) showed that, even when just 60-89 year olds were used, a significant age effect existed in performance on the TOL. By breaking this range into 3 separate groups, the time at which these changes are most significant can be more precisely investigated.

When creating a computerized test, it is not always a process of simply taking the exact same test and putting it into a digital format. When a test is converted, there is a chance that the outcome measures of the manual and computer forms of a test may represent different constructs and tap into different abilities. The present study will utilize the same database from which the TOLDX data was drawn to see how areas that have been correlated with manual TOL performance, WM, fluid intelligence, and age, relate to performance on a computerized version of the same test.

Participants The majority of participants will be drawn from an active database currently consisting of over 7,200 individuals. The participants in the present study will range in age from 16 to 89 years of age. The lower bound of this age range is based on the lower age cutoff on the adult version of the Tower of London-Drexel Second Edition (TOLDX). The participants will be divided into 8 different groups based on their age; the groups will be divided into age groups 16-19, 20s, 30s, 40s, 50s, 60s, 70s, and 80s. The mean age, years of education, and the proportion of males and females for each group are shown in Table 1.

Table 1

Participants were and will be recruited from the University of Colorado at Colorado Springs psychology classes; older adults will be recruited from the Gerontology Center database; 16 and 17 year old participants, as well as additional older adult participants, will be recruited from relatives of the student and older adult participants.

Participants will be excluded from participation in the study if they report any current or past neurological disease (schizophrenia, dementia, etc.), traumatic brain injury, learning disability, major psychiatric condition (depression, generalized anxiety disorder, etc.) or are currently using a substance they believe has an effect on their cognition; either illicit or prescribed (Ritalin, cannabis, etc.).

Participants who are 16 or 17 years of age will receive $5 per hour for their participation. College students will receive extra credit or SONA research credits for each hour of participation. SONA credits are awarded to students for participation in campus based research. Many undergraduate psychology courses require a minimum number SONA credits as a component of the students overall grade. Older adults (age 60 and older) will receive $10 for each hour of participation. Each testing session will be limited to a maximum of two hours per day. Testing sessions will be capped at two hours to reduce any variance that may arise due to participant fatigue.

Instruments  Prior to participation in the study, all participants will complete an extensive demographic and general health questionnaire that will provide further information on the general health of participants.

Participants will be administered a computerized version of the TOLDX in a quiet testing room. The computerized TOLDX is presented on a desktop computer and can be manipulated using either a mouse or touch screen. A computerized schematic of a typical problem is shown in Figure 1.

FIGURE 1 MARKER

The computerized TOLDX is comprised of 10 trials following three practice trials that range in minimum number of moves from 4 to 7 and participants are given a maximum of 20 moves and 2 minutes within which to complete the trial. Prior to beginning the test, the rules are both read, using a digitized voice, and displayed below a practice problem. The rules are read as follows:

See these two boards? They are both alike. This board on the left is the one you will use and you will make it look like the one on the right. Your task is to make this arrangement on the left look like the one on the right in as few moves as possible. There are two rules you must follow when you are arranging the beads. The first rule is that you are not allowed to place more beads on a peg than it will hold. The second rule is that you can only move one bead at a time. You cannot move two beads off the pegs at the same time. Do you have any questions? Now, arrange the beads on the left so they look like the arrangement on the right. You have two minutes to do each problem. Also, you must complete the problem in 20 moves or fewer. If you do not finish in two minutes or in less than 20 moves, the trial ends and a new problem is presented.

Research Design  Following the initial information gathering, all participants will complete a variety of cognitive and neuropsychological tests. Each one hour session of testing will include a variety of tests that assess memory, problem solving, inhibition and attention, task switching ability, and general intelligence. In addition to the manual TOLDX, there are two other versions of the TOL that the participants could take; an identical manual version of the TOLDX and another computerized version comprised of thirty-one unique trials and a shorter, 60 second time limit. On any given visit, no participant will be given more than one version of the TOL and, if they choose to return, there will be a minimum of two days between testing to reduce potential practice effects.

When completing the TOLDX the participant will be brought into a quiet room and seated in front of a computer. The previously stated rules will be presented on the computer prior to beginning the test and any questions will be answered. The TOLDX is a combination of a time and performance limited test in which they are allowed a maximum of 20 moves or 2 minutes per trial. If they fail to complete the trial within the set parameters, the time and total number of moves will be recorded as the highest possible value (120 seconds and 20 total moves – minimal number of moves for perfect solution) and they move to the next trial. The number of excess moves for problems in which the participant exceeds the allotted time or number of moves will reflect the minimum number of moves possible for that problem subtracted from 20. The participants will be presented with three practice problems to learn the rules of the TOLDX, followed by the ten trials that make up the test itself.

All timing and recording of moves will be recorded digitally. The dependent measures that will be taken into account in this study are the initial/think time, the total time, the execution time which can be calculated by subtracting the initial time from the total time, and the total number of excess moves. Total perfect solutions.Each of these measures will be recorded for each trial. The independent variable that will be assessed for these measures will be the participant’s age condition (their preset age range).

Results  The present study will be investigating the effect of age, sex, years of education and fluid intelligence as assessed by performance on the Matrices subtest of the Wechsler Abbreviated Scale of Intelligence-1st edition (WASI) on total excess moves, total initial time, and total time. The time and move values will be totaled across all 10 trials of the TOLDX. All analyses will be run through SPSS.

A multiple regression will be run for each of the outcome measures; excess moves, total initial time, and total completion time. The effect of age, sex, years of education, and Matrices raw score will be assessed to detect the level of variance contributed by each condition. Pearson regression analyses will also be run across these factors to detect any effects that may have been lost due to shared variance. It is hypothesized in the multiple regression that all factors will have an effect on excess moves, initial time, and total time with Matrix score and age having the greatest effect, and years of education and sex having the smallest effects. Similar results are hypothesized for the regression analyses.

Lastly, univariate analyses will be run to detect the effect of age condition (20s, 30s, etc.) on total excess moves, total initial time, and total completion time. It is hypothesized that a U shape will exist across all three outcomes, such that excess moves and total time will be highest in adolescents and older adults with peak performance occurring in early adulthood around during the 30s. Initial times are hypothesized to be shortest in adolescents and older adults and longest in early adulthood.

References

Albert, D. & Steinberg, L. (2011). Age differences in strategic planning as indexed by the tower of london. Child Development, 82(5), 1501-1517. Alloway, T.P. & Alloway, R.G. (2013). Working memory across the lifespan: A cross-sectional approach. Journal of Cognitive Psychology, 25(1), 84-93. doi: 10.1080/20445911.2012.748027 Andrés, P., & van der Linden, M. (2000). Age-related differences in supervisory attentional system functions. Journals of Gerontology, Series B Psychological Sciences and Social Sciences, 55, P373-P380. doi: 10.1093/geronb/55.6P373 Arbuthnott, K. & France, J. (2000). Trail making test, part b as measure of executive control. Journal of Clinical and Experimental Neuropsychology, 22, 518-528. Baddeley, A.D. (1996). Exploring the central executive. Quarterly Journal of Experimental Psychology, 49A, 59-28. Berg, W.K. & Byrd, D.L. (2002). The tower of london spatial problem-solving task: Enhancing clinical and research implementation. Journal of Clinical and Experimental Neuropsychology, 24(5), 586-604. Berg, W.K., Byrd, D.L., McNamara J.P.H., & Case, K. (2010). Deconstructing the tower: Parameters and predictors of problem difficulty on the tower of london task. Brain and Cognition, 72, 472-482. Berg, W.K., Byrd, D.L., McNamara, J.P.H., & MacDonald, C.A. (2006). New indices of planning abilities using the tower of London task. The Quantified Process Approach to Neuropsychological Assessment, 225-248. Bishop, D.V., Aamodt-Leeper, G., Creswell, C., McGurk, R., & Skuse, D.H. (2001). Individual differences in cognitive planning on the tower of hanoi task: Neuropsychological maturity or measurement error? Journal of Child Psychology and Psychiatry, 42, 551-556. Bottari, C., Dassa, C., Rainville, C., & Dutil, E. (2009). The criterion-related validity of the IADL profile with measures of executive functions, indices of trauma severity and sociodemographic characteristics. Brain Injury, 23(4), 322-335. http://dx.doi.org/10.1080/02699050902788436 Brundson, R., Nickels, L., & Coltheart. (2007). Topographical disorientation: Towards an integrated framework for assessment. Neuropsychological Rehabilitation, 17, 34-52. Bugg, J.M., Zook, N.A., DeLosh, E.L., Davalos, D.B., & Davis, H.P. (2006). Age differences in fluid intelligence: Contributions of general slowing and frontal decline. Brain and Cognition, 62, 9-16. doi: 10.1016/j.bandc.2006.02.006 Carlin, D., Bonerba, J., Phipps, M., Alexander, G., Shapiro, M., & Grafman, J. (2000). Planning impairments in frontal lobe dementia and frontal lobe lesion patients. Neuropsychologia, 38, 655-665. http://dx.doi.org/10.1016/S0028-3932(99)00102-5 Cespón, J., Galdo-Álvarez, S., & Díaz, F. (2013). Age-related changes in ERP correlates of visuospatial and motor processes. Psychophysiology, 50, 743-757. doi: 10.1111/psyp.12063 Cheetham, J.M., Rahm, B., Kaller, C.P., & Unterrainer, J.M. (2012). Visuospatial over verbal demands in predicting tower of london planning tasks. British Journal of Psychology, 103, 98-116. Cowan, N., Naveh-Benjamin, M., Kilb, A., & Saults, J. S. (2006). Lifespan development of visual working memory: When is feature binding difficult? Developmental Psychology, 42, 1089–1102. doi:10.1037/0012-1649.42.6.1089 Culbertson, W.C., Moberg, P.J., Duda, J.E., Stern, M.B., & Weintraub, D. (2004). Assessing the executive function deficits in patients with parkinson’s disease: Utility of the tower of London-drexel. Assessment, 11(1), 27-39. doi: 10.1177?1073191103258590 Culbertson, W.C. & Zillmer, E.A. (1998a). The construct validity of the tower of londondx as a measure of the executive functioning of ADHD children. Assessment, 5, 215-226. Culbertson, W.C. & Zillmer, E.A. (1998b). The tower of londonDX: A standardized approach to assessing the executive functioning in children. Archives of Clinical Neuropsychology, 13, 285-301. Culbertson, W.C. & Zillmer, E.A. (1998c). The Tower of LondonDX Manual. Multi-Health Systems, Inc. Culbertson, W.C. & Zillmer, E.A. (2001). Tower of london-drexel university (TOL¬DX): 2nd edition technical manual. North Tonawanda, NY: Multi-Health Systems Inc. Debelak, R., Egle, J., Kötering, L., & Kaller, C.P. (2015) Assessment of planning ability: Psychometric Analyses on the unidimensionality and construct validity of the tower of London task (TOL-F). Neuropsychology. Advance online publication. http://dx.doi.org/10.1037/neu000238 Diamond, A (2013). Executive functions. Annual Review of Pscyhology, 64, 135-168. http://dx.doi.org/10.1146/annurev-psych-113011-143750 Embertson, S.E. (1983). Construct validity: Construct representation versus nomothetic span. Psychological Bulletin, 93, 179-197. Embertson, S.E. (1998). A cognitive design system approach to generating valid tests: Application to abstract reasoning. Psychological Methods, 3, 380-396. http://dx.doi.org/10.1037/1082-989X.3.3.380 Fandakova, Y., Sander, M.C., Werkle-Bergner, M., & Shing, Y.L. (2014). Age differences in short-term memory binding are related to working memory performance across the lifespan. Psychology and Aging, 29(1), 140-149. Finkel, B., Reinolds, C.A., McArdle, J.J., & Pederson, N.L. (2003). Latent growth curve analyses of accelerating decline in cognitive abilities in late adulthood. Developmental Psychology, 39, 533-550. doi: 10.1037/0012-1649.39.3.535 Gates, N.J. & Kochan, N.A. (2015). Computerized and on-line neuropsychological testing for late-life cognition and neurocognitive disorders: Are we there yet? Current Opinion Psychiatry, 28, 166-172. doi: 10.1097/YCO.0000000000000141 Gilbert, S. & Shallice, T. (2002). Task switching. Cognitive Psychology, 44, 297-337. Gilhooly, K.J., Wynn, V., Phillips, L.H., Della Sala, S. (2002). Visuospatial and verbal working memory in the five disc tower of london task: An individual differences approach. Thinking and Reasoning, 8, 165-178. Hayes-Roth, B. & Hayes-Roth, F. (1979). A cognitive model of planning. Cognitive Science, 3, 275-310. Hedden, T., & Gabriele, J.D.E. (2004). Insights into the ageing mind: A view from cognitive neuroscience. Nature Reviews Neuroscience, 5, 87-96. doi: 10.1038/nrn1323 Jacobs, R. & Anderson, V. (2002). Planning and problem solving skills following focal frontal brain lesions in childhood: Analysis using the tower of london. Child Neuropsychology, 8, 93-106. http://dx.doi.org/10.1076.chin.8.2.93.8726 Joyce, E.M. & Robbins, T.W. (1991). Frontal love function in korsakoff and non-korsakoff alcoholics: Planning and spatial working memory. Neuropsychologia, 29(8), 709-723. Kaller, C.P., Rahm, B., Köstering, L., & Unterrainer, J.M. (2011). Reviewing the impact of problem structure on planning: A software tool for analyzing tower tasks. Behavioural Brain Research, 216, 1-8. doi: 10.1016/j.bbr.2010.07.029 Kaller, C.P., Unterrainer, J.M., Rahm, B., & Halsband, U. (2004). The impacts of problem structure on planning: Insights from the tower of london task. Cognitive Brain Research, 20, 462-472. (11) http://dx.doi.org/10.1016/j.cogbrainres.2004.04.002 Kaller, C.P., Unterrainer, J.M., & Stahl, C. (2012). Assessing planning ability with the tower of london task: Psychometric properties of a structurally balanced problem set. Psychological Assessment, 25(1), 46-53. Köstering, L., Schmidt, C.S.M., Egger, K., Amtage, F., Peter, J., Klöppel, S., Beume, L-A., Hoeren, M., Weiller, C., & Kaller, C.P. (2015). Assessment of planning performance in clinical samples: Reliability and validity of the tower of london task (TOL-F). Neuropsychologia, 75, 646-655. Köstering, L., Stahl, C., Leonhart, R., Weiller, C., & Kaller, C.P. (2014). Development of planning abilities in normal aging: Differential effects of specific cognitive demands. Developmental Psychology, 50(1), 293-303. doi: 10.1037/a0032467 Krishnan, M., Smith, N., & Donders, J. (2012). Use of the tower of london-drexel university, second edition (TOLDX) in adults with traumatic brain injury. The Clinical Neuropsychologist, 26(6), 951-964. doi: 10.1080/13854046.2012.708166 Larochette, A.C., Benn, K., & Harrison, A.G. (2009). Executive functioning: A comparison of the tower of londonDX and the d-kefs tower test. Applied Neuropsychology, 16, 275-280. doi: 10.1080-09084280903098695 Masson, J.D., Dagnan, D., & Evans, J. (2010). Adaptation and validation of the tower of london test of planning and problem solving in people with intellectual disabilities. Journal of Intellectual Disability Research, 54(5), 457-467. McArdle, J.J., Ferrer-Caja, E., Hamagami, F. & Woodcock, R.W. (2002). Comparative longitudinal structural analyses of the growth and decline of multiple intellectual abilities over the life span. Developmental Psychology, 38, 115-142. doi: 10.1037/0012-1649.38.1.115 Morice, R. & Delahunty, A. (1996). Frontal/executive impairments in schizophrenia. Schizophrenia Bulletin, 22(1), 125-137. Moscovitch, M., & Winocur, G. (1995). Frontal lobes, memory, and ageing. In J. Grafman, K.J. Holyoak, & F. Boller (Eds.), Annals of the New York Academy of Sciences: Vol. 769, Structure and functions of the human prefrontal cortex (pp. 119-150). New York, NY: New York Academy of Sciences. Newman, S.D. & Pittman, G. (2007). The tower of london: A study of the effect of problem structure on planning. Journal of Clinical and Experimental Psychology, 29, 333-342. Owen, A.M. (2005). Cognitive planning in humans: New insights from the tower of London (TOL) task. The Cognitive Psychology of Planning, 135-151. Pulos, S. & Denzine, G. (2005). Individual differences in planning behavior and working memory: A study of the tower of london. Individual Differences Research, 3(2), 99-104. Shallice, T. (1982). Specific impairments in planning. Philosophical Transactions of the Royal Society B, 298, 199-209. Palermo, L., Piccardi, L., Nori, R., Giusberti, F., & Guariglia, C. (2016). The impact of ageing and gender on visual mental imagery processes: A study of performance on tasks from the complete visual mental imagery batter (CVMIB). Journal of Clinical and Experimental Neuropsychology, 38(7), 752-763. doi: 10.1080/13803395.2016.1161735 Riccio, C.A., Wolfe, M.E., Romine, C., Davis, B., & Sullivan, J.R. (2004). The tower of London and neuropsychological assessment of adhd in adults. Department of Educational Psychology, 19, 661-671. doi: 10.1016/j.acn.2003.09.001 Salthouse, T.A. (1985). Speed of behavior and its implications for cognition. In J.E.Birren & K.W. (Eds.), Handbook of psychology of aging (pp. 400-426). New York, NY: Van Nostrand Reinold. Shallice, T. & Burgess, P. (1991). Higher-order cognitive impairments and frontal lobe lesions in man. In H.S. Levin, H.M. Eisenberg, & A.L. Benton (Eds.), Frontal lobe function and dysfunction (pp. 125-138). New York: Oxford University Press. Steinberg, L., Albert, D., Cauffman, E., Banich, M., Graham, S., & Woolward, J. (2008). Age differences in sensation seeking and impulsivity as indexed by behavior and self-report: Evidence for dual systems model. Developmental Psychology, 44, 1764-1777. Towse, J.N., Hitch, G.J., & Hutton, U. (2002). On the nature of the relationship between processing activity and item retention in children. Journal of Experimental Child Psychology, 82, 156-184. Unterrainer, J.M. & Owen, A.M. (2006). Cognitive executive function: From neuropsychology to neuroimaging. Journal of Physiology, 99, 308-317. Ward, G., & Allport, A. (1997). Planning and problem-solving using the five-disc tower of london task. The Quarterly Journal of Experimental Psychology, 50A, 49-78. Zygouris, S. & Tsolaki, Magda (2015). Computerized cognitive testing for older adults: A review. American Journal of Alzheimer’s Disease & Other Dementias, 30(1), 13-28.