Intra Individual Variability in Ageing and Cognition

Print   

03 Apr 2018

Disclaimer:
This essay has been written and submitted by students and is not an example of our work. Please click this link to view samples of our professional work witten by our professional essay writers. Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of EssayCompany.

When investigating cognitive functioning in adulthood, research has typically been focused on average age-related effects. In terms of research methodology, this has translated into the comparison of mean level performance in different age groups using either longitudinal or cross sectional designs. This research has been useful, but it is based on the assumption that any investigated behaviours are stable across time or that any trajectory of change observed is similar for all persons (Hultsch et al., 2011). This assumption represents an instance of a more general stability perspective which has been predominant in developmental research (Gergen, 1977, Nesselroade & Featherman, 1997). This stability perspective attributes short term fluctuations in performance to error, therefore failing to adequately explain them. In order to demonstrate that short term fluctuations in cognitive functioning during ageing have validity in their own right, two conditions must be met. Firstly, it is necessary that these short term fluctuations in performance can be reliably measured. Second, it must be shown that short term fluctuations in performance can allow insights into cognitive functioning during aging that average age-related effects cannot.

These short term fluctuations in performance are referred to as Intra individual variability. Intra individual variability refers to reversible fluctuations in behaviour over short periods of time (Hultsch et al., 2011). Three key features of intra individual reliability can be drawn from this definition. Firstly, intra individual variability is concerned with the variability of constructs (e.g. behaviour or performance) within the same person (within-person differences) as opposed to variability which differentiates individuals from each other (between-person differences). Second, intra individual reliability is suggested to be a rapid phenomenon with potentially observable cycles involved as a result of its occurrence over short time periods. Lastly, even though the short term fluctuations observed in intra individual reliability are temporary and reversible, they are reliable within a given individual (Stawski et al., 2014).

Reliability and Magnitude of Intra individual reliability

Although there are certain types of intra individual variability which appear to follow some sort of cyclical rhythm (e.g. alertness related circadian rhythms (Webb, 1982)), much within person variability appears to be “random” in nature. The fact that no cyclical or other patterns can be detected within the variability raises an important question. Is the phenomenon of intra individual variability legitimate for study? The traditional stability perspective of psychological measurement would assume that any observed score is composed of a true score plus error. In other words, any intra individual variability observed is typically treated as error (Hultsch et al., 2011).

Evidence has been found for the existence of substantial amounts of intra individual variability in what have previously been considered to relatively stable, trait like cognitive functions. This has been achieved through the evaluation of the magnitude of intra individual reliability on a certain task as compared to the magnitude of between person variability on the same task. By utilising this approach, many studies have reported that the amount of intra individual variability is about half the amount of between person variability on the same task (Li et al., 2001a; Nesselroade & Salthouse, 2004; Salthouse et al., 2006). For instance, Nesselroade & Salthouse (2004) asked participants (20-91 years) to complete three perceptual motor tasks at three separate times over a two week period. Three measures of variability were computed for each task as follows: (a) within person variability (standard deviation of participants three mean scores for the three sessions; (b) within person variability (average of participant’s trial to trial standard deviations for all sessions; (c) between person variability (standard deviation of mean scores of participants). The ratio of both types of within person variability to between person variability was examined which revealed that the majority of values ranged from .40 to .55. Hence, it was found that intra individual variability was about half of inter individual variability in the perceptual motor tasks. Salthouse et al., (2006) found similar results for variability in accuracy of participants (18-97 years) on multiple cognitive tasks over a period of two weeks. Both of these studies used a wide age range but analogous results have also been reported using a much smaller adult age range (64-86 years). Lin et al., (2001a) investigated participants intra individual variability across a group of sensorimotor and memory variables in 13 fortnightly sessions. Uniform with the previous two studies, the magnitude of the intra individual variability was half that of the inter individual variability in both variables in the sample. Taken together, these finding suggest that there is a significant amount of intra individual variability found in traits which are traditionally considered to be stable in nature. Furthermore, it is a point of interest that the magnitude of the intra individual variability seems to be quite similar across different age ranges and cognitive tasks (Hultsch et al., 2011).

The question still remains of whether intra individual variability could simply be characterised as error and not as a legitimate psychological phenomenon. There are two ways in which to view intra individual variability which are not necessarily mutually exclusive. The first is to regard the variability as “noise which should be dealt with in order to ameliorate any measured “signal”. Alternatively, intra individual variability may be viewed as a “signal” within its own right and hence attempts must be made to investigate it (Nesselroade & Salthouse, 2004). One line of evidence to support the latter view comes from experiments which have used multiple observations to create a time series. For instance, spectral analysis forms a frequency domain from the time domain in order to examine the data for periodicity. If the series is random, all frequencies should be equally prominent which would indicate “white noise”. However, if the series is periodic in nature, a peak should be shown at the “driving frequency” in the resulting power spectrum, indicating the presence of systematic variation (coloured or pink noise). Time series analyses generated using data from several tasks including choice reaction time tasks (Clayton & Bruhns Frey, 1997) and time estimation (Gilden et al., 1995) have reported that trial to trial fluctuations in these tasks are incongruent with the view that intra individual variability is simply due to a large amount of independently occurring random variables (“white noise”). Instead, spectrum analysis proposes two components of intra individual variability. Namely, a coloured noise component and a white noise component, which is indicative of the lawful variation being present in the data (Hultsch et al., 2011).

The argument that intra individual variability can be measured reliably gains further evidence from studies which examine inter-correlations of variability measures which are derived from different tasks and/or collected on different occasions (Hultsch et al., 2011). Several studies have reported that the magnitude of intra individual variability seems to be unique to the individual. A number of studies have shown two results which are in line with this view (e.g. Fuentes et al., 2001; Hultsch et al., 2002; Rabbit et al., 2001). Firstly, intra individual variability measures in a given task taken across different time intervals (e.g. later versus earlier sessions) are correlated positively (Allaire & Marsiske, 2005; Hultsch et al., 2000; Rabbit et al., 2001). Additionally, in reaction time taks, greater intra individual variability on one task correlates with greater intra individual variability in other tasks in the same individuals (Fuentes et al., 2001; Hultsch et al., 2000; 2002). Comparable results which show stability in participant’s intra individual variability over time have been reported in studies which have investigated different functional domains such as cognitive and physical functioning (Strauss et al., 2002).

Homogenous with the relationships described above, it has been indicated by studies which have analysed reliability between odd and even trials that intra individual variability measures have good reliability (Hultsch et al., 2011). For example, Sliwinski et al., (2006) analysed the reliabilities of numerous reaction time tasks based on each participants even and odd trials. Acceptable reliability levels were found which ranged from .65 to .78 across separate age groups and tasks.

Taken together, the research outlined above shows that intra individual variability is not only reliable as a measure of cognitive functions, but also has a high magnitude. However, it could be argued that the most prominent evidence for the reliability of intra individual reliability measures comes from their systematic association with performance outcomes and personal characteristics, including age and cognitive functioning. Hence, this association will presently be explored in further depth.

Intra individual variability in ageing and cognition

If it is accepted that intra individual variability in performance is a meaningful index of human behaviour, the possibility ensues that it may be useful to further investigate this variability in order to gain further insights into psychological phenomenon. For several reasons, it follows that intra individual variability could be of particular interest to cognitive psychologists interested in ageing Hultsch et al, 2011). Firstly, cognitive ageing is strongly indicated by intra individual variability. More specifically, intra individual variability is thought to be an index of the functioning of the central nervous system (Dykiert et al., 2012). For instance, Hendriickson (1982) suggested that the cause of reaction time intra individual variablilty may be random errors (neural “noise) in central nervous system signal transmission. More recent theories by Li et al., (1999; 2001) have suggested that neural “noise” may be modulated by various neurotransmitters including norepinephrine and dopamine. Behavioural studies provide support this notion as many studies have observed greater reaction time intra individual variability in a number of different conditions which affect the functioning of the centeral nervous system (Dykiert et al., 2012). For example, following a traumatic brain injury (Stuss et al., 1989) in neurodegenerative diseases such as Alzheimer’s, Parkinson’s disease (Burton et al., 2006) or dementia (Gordon & Carson, 1990). Recation time intra individual variability has also been found to be higher in states which cause a temporary loss of functioning of the central nervous system such as presence at high altitude (McFarland, 1937a; 1937b) and the consumption of alcohol (Maruff et al., 2005). As a whole, this evidence suggests that the deterioration of the central nervous system is marked by greater intra individual variability in reaction time. These findings have also led certain researchers to suggest that intra individual variability can provide a unique insight into cognitive functioning beyond the scope of average age related effects (Hultsch et al., 2005, Hultsch & MacDonald, 2004; Rabbit et al., 2001).

If intra individual variability in reaction time is an index of central nervous system functioning, it would be expected that older adults would display greater intra individual variability than younger adults. Indeed, many studies have found a significant differences between age groups in reaction time intra individual variability. A plethora of studies have reported greater reaction time intra individual variability in older adults when compared to younger adults in various reaction time tasks (e.g. Antsey, 1999; Fozard et al., 1994; Hultsch et al., 2002; Nesselroade & Salthouse, 2004; West et al., 2002). Hultsch et al. (2002) have suggested that age differences in reaction time intra individual variability within the older age group can also be observed, especially after approximately age 75.

This general pattern of age effects is demonstrated by cross sectional results reported by Hultsch et al. (2002). Age differences in reaction time intra individual variability were examined in younger adults (19-36 years) and three separate groups of older adults (young-old, 54-64 years; mid-old, 65-74 years; old-old, 75-94 years). Two verbal and two non-verbal reaction time tasks were completed by all participants and consequent intra individual standard deviations were computed trial by trial in each task. Greater intra individual variability was observed in the three older age groups in all tasks when compared to the younger adults. Effect sizes of age group differences typically ranged from medium to large.

Method

Participants

Results

Two Choice Reaction Time Task (RT IIV)

Table 1

A table to show the variability in participant’s reaction times (Reaction Time Intra individual Standard Deviation) during the Two Choice Reaction Time task.

Age

Mean

SD

Young

5.507

2.942

Younger Middle Aged

6.398

3.104

Older Middle Aged

8.272

5.509

Old

8.632

2.528

A one-way between subjects ANOVA was conducted to compare the effect of age (young, younger middle-aged, older middle-aged, old) on RT ISD (reaction time intra-individual standard devation) in the Two Choice Reaction Time task. There was a significant effect of age on RT ISD at the p<.05 level for the four age groups (F(3,103) = 4.653, p = .004). Hochbergs GT2 post-hoc comparisons were applied to the data to test for significant differences between each age group. Hochbergs GT2 was selected as the group sample size for the young group was larger (N = 42) than the younger middle-aged (N = 23), older middle-aged (N = 24) or old (N = 18) groups. Post hoc comparisons indicated that the mean RT ISD for the young group was significantly different than both the older middle-aged and the old group. However, the mean RT ISD for the younger middle-aged group did not significantly differ from any of the other age groups. Additionally, the mean RT ISD’s for the older middle-aged and the old group did not differ significantly.

Table 2

A table to show the variability in participant’s reaction times (Reaction Time Covariance) during the Two Choice Reaction Time task.

Age

Mean

SD

Young

0.181

0.062

Younger Middle Aged

   

Older Middle Aged

0.224

0.075

Old

   

An additional ANOVA was conducted to compare the effect of age on RT CV (reaction time covariance) in the Two Choice Reaction Time Task. A significant effect of age was found at the p<.05 level for the four age groups [F(3, 105) = 3.014, p = .33]. Hochbergs GT2 post-hoc comparisons were applied to the data to test for significant differences between the four age groups. Post-hoc comparisons indicated that the mean RT CV for the young group was significantly different than the older middle-aged group. However, no other age groups were found to be significantly different from one another.

Taken together, these results suggest that there is a significant effect of age on RT IIV in the Two Choice Reaction Time Task.

Touch Screen Task (RT IIV)

Table 3

A table to show the variability in participant’s reaction times (Reaction Time Intra individual Standard Deviation) during the Touch Screen task.

Age

Mean

SD

Young

   

Younger Middle Aged

   

Older Middle Aged

   

Old

   

A one-way between subjects ANOVA was conducted to compare the effect of age (young, younger middle-aged, older middle-aged, old) on RT ISD in the Touch Screen Task. There was no significant effect of age on RT ISD at the p<.05 level [F(3, 102) = 1.504, p = .218]. No post-hoc comparisons were applied as no significant effect was found. These results suggest that age does not have a significant effect on the RT IIV of participants in a touch screen task.

Table 4

A table to show the variability in participant’s reaction times (Reaction Time Intra individual Standard Deviation) during the Touch Screen task.

Age

Mean

SD

Young

   

Younger Middle Aged

   

Older Middle Aged

   

Old

   

A further one-way ANOVA was conducted to compare the effect of age on RT CV in the Touch Screen Task. No significant effect of age was found at the p<.05 level for the four age groups [F(3, 106) = .732, p = .535]. No post-hoc comparisons were applied as no significant effect was found.

Taken together, these results indicate that there is no significant effect of age on RT IIV in the touch screen task

Touch Screen Task (Acc IIV)

Table 5

A table to show the variability in participant’s accuracy (accuracy raw standard deviation) during the Two Choice Reaction Time task.

Age

Mean

SD

Young

   

Younger Middle Aged

   

Older Middle Aged

   

Old

   

A one way between subjects ANOVA was conducted to compare the effect of age (young, younger middle-aged, older middle aged, old) on Acc SD (Accuracy raw standard deviation) in the Touch Screen Task. No significant effect of age on Acc SD was found for the four age groups at the p<.05 level [F(3, 106) = 1.529, p = .211]. No post-hoc comparisons were conducted as no significant effect was found. The results suggest that age does not have an effect on the intra individual variability of the accuracy of participants in a choice reaction time touch screen task.



rev

Our Service Portfolio

jb

Want To Place An Order Quickly?

Then shoot us a message on Whatsapp, WeChat or Gmail. We are available 24/7 to assist you.

whatsapp

Do not panic, you are at the right place

jb

Visit Our essay writting help page to get all the details and guidence on availing our assiatance service.

Get 20% Discount, Now
£19 £14/ Per Page
14 days delivery time

Our writting assistance service is undoubtedly one of the most affordable writting assistance services and we have highly qualified professionls to help you with your work. So what are you waiting for, click below to order now.

Get An Instant Quote

ORDER TODAY!

Our experts are ready to assist you, call us to get a free quote or order now to get succeed in your academics writing.

Get a Free Quote Order Now