The Influence of Acquiring a Second Language
on Executive Functioning and the mediating role
of Individual Differences in Elderly
Olgers, B. E. 10588841
Bachelorproject Brain & Cognition University of Amsterdam
Supervisor: J. Buitenweg May 27th 2016
Abstract
Normal aging is associated with a decline of executive functions. This has a negative association with overall quality of life. In addition, individuals vary a lot in the degree of this decline. However, bilingualism in elderly is associated with a slower decline of executive functions. Acquiring a second language at an older age might have this same effect. This study tried to investigate the effect of acquiring a second language on the executive functions inhibition, switching and updating and the mediating role of the individual differences: intelligence, quality of sleep and physical activity in elderly.
Thirteen participants completed a three-week Italian language course. Before and after taking this course, their switching, inhibition and updating ability was measured. In addition, their intelligence was measures and their average quality of sleep and physical activity.
The study revealed no improvement of executive functioning after learning a second language and also no influence of the individual differences mediating this relationship. However, the participants did not have enough time to acquire the second language and this could be part of the reason that no effect was found. For future studies, the recommendations of this research should be considered. Decline of executive functioning is a universal aspect of aging, and because of its negative association with quality of life, research trying to slow down this decline is of great value.
Table of contents
Introduction………..p. 4 Method……….p. 9 Subjects……….p. 9 Measures………..p. 10 Procedure……….p. 13 Results………p. 13 Discussion………p. 21 References………p. 29Introduction
Cognitive decline is a universal aspect of the aging process. Most elderly experience non-pathological losses in their cognitive functioning. This might not always profoundly affect their real-world functioning, but still even a modest reduction in cognitive function is negatively associated with quality of life, independence, frequency and quality of social interaction and engagement in cognitively stimulating activities (Mahncke et al., 2006). In addition, the lifespan of the world’s aging population is extending dramatically; people are getting older and older. However, the human cognitive span does not match this
extended lifespan yet. Several longitudinal studies have shown that, on average, a decline of cognition can be found in in old age (Wilson et al., 2002). Especially executive functions are prone to this decline associated with aging (Treitz et al., 2007). Yet, on an individual level, a lot of variability can be found. Some elderly experience extensive cognitive decline, while others only experience a slight decline or not even a one at all. Knowledge about which factors contribute to this variability is, however, limited (Wilson et al., 2002). Therefore research should be conducted, to not only determine if the human cognitive span can be matched more closely to this aging population with increased cognitive decline, but also consider the individual differences playing a roll in this (Chapman et al., 2015).
An interesting first question to ask is if cognitive decline is a process that can be reduced, or that it is just a consequence of living longer. To examine if this decline could be reduced, the plasticity, the functional and structural changes in the brain, should be explored to determine to what extend an older brain is still capable of such plasticity (Chapman et al., 2015). Even with normal aging, dramatic changes in brain networks occur. Atrophy for grey, as well as white
matter, and loss of synaptic connections have been reported (Schlee et al. 2012). Chapman et al. (2015) studied changes in an aging brain after mental training. They found that cognitive training enhances resting-state neural activity and connectivity, increasing the blood supply to these regions. This gives evidence that even in old age, the brain still shows plasticity and that cognitive training might be useful to prevent or slow down cognitive decline in elderly.
So, cognitive training could possibly slow down the cognitive decline experienced in elderly. Previous research already attempted to achieve this. For instance, Cheng et al. (2012) found that cognitive training can improve memory, visual reasoning, visuo-spatial construction, and attention. In addition, a review by Lustig et al. (2009) reported that numerous studies of cognitive training in older adults found substantial improvement in the trained task and near transfer tasks. However, modest to no effects on far transfer task or daily activities were found. This lack of far transfer was also reported by Ball et al. (2002). In their research on the effect of cognitive training on everyday functioning, they did find an effect of their training on the trained aspects, but they did not find any training effects on everyday functioning. The reason for this lack of far transfer might be because some of the trainings used are not really applicable to real-life tasks (Lustig et al., 2009).
Thus, evidence suggests that cognitive training does have a positive effect on the specific trained cognitive traits and on similar tasks, but a training that has an effect on daily functioning is not yet found. There might be a need for a training that would be more similar to a real life situation. For instance, learning a second language might be able to enhance far transfer (Buitenweg et al., 2012).
Evidence has shown that lifelong bilingualism can slow down age-related cognitive decline, especially in executive functioning. Current models suggest that this advantage may be the result of bilinguals managing two languages at all times, which requires executive resources like selecting the right language, inhibit the language not in use, and be able to switch between the languages. (Bialystok & Craik, 2010; Soveri et al., 2011). These executive functions that seem to be most influenced by bilingualism are also the functions first affected by normal aging (Treitz et al., 2007). Executive functions are often divided in three separate domains: switching, updating and inhibition (Miyake et al., 2000).
Switching is a process that involves a person’s ability to switch their attention and actions between relevant tasks (Buitenweg et al., 2012). Bilinguals have to constantly switch between two languages, which probably strengthens their switching ability. This was confirmed by Gold et al. (2013), who found that bilingual elderly performed better at a perceptual switching task than their monolingual peers.
Inhibition is as a process that requires selective attention, for instance inhibiting inappropriate responses (Kramer & Mota, 2015). Bilinguals have to suppress their non-relevant language at all times, which gives them a massive practice at inhibition processes. Proof for this theory came from Bialystok et al. (2004) who found better inhibitory control in older bilinguals compared to older monolinguals, and this bilingual advantage increased with age. Also, Kramer & Mota (2015) found that late bilinguals (who spoke a second language for at least two months) showed better inhibitory control than monolinguals. So it seems that this bilingual advantage on inhibition tasks is not only restricted to a
life-long bilingualism, but can also be acquired by learning a second language later on.
Updating is a process consisting of monitoring incoming information for its relevance and adapting the content of working memory storage accordingly (Buitenweg et al., 2012). Bilinguals have to monitor the environment to ensure use of the appropriate language. Luo et al. (2013) found that bilinguals
outperformed monolinguals on a simple span task, that required updating of information.
Thus, with growing age people experience more cognitive decline, especially in executive functioning (Treitz et al. 2007). However, bilinguals outperform monolinguals on all these executive functions (Gold et al., 2013; Bialystok et al., 2004; Luo et al., 2013), so bilingualism might act as a preserver of the switching, inhibition and updating ability. Therefore these advantages might also be established in monolinguals when they acquire a second language. In addition, it would be interesting to look at individual differences mediating the relationship between learning a second language and executive functioning in elderly, because a lot of variability on the individual level of cognitive
functioning can be found (Wilson et al., 2002).
The first individual difference that will be looked at is intelligence. A lot of studies have already shown that higher levels of intelligence are good predictors of which individuals can sustain greater brain damage before
experiencing cognitive decline. These individuals are expected to have a greater ‘cognitive reserve’, which means that they can use their neural networks more efficiently (Stern, 2002). In addition, Murayama et al. (2013) found that full scale IQ in elderly had a stronger correlation with memory function than years of
education. Intelligence thus seems to be an important predictor of cognitive decline in elderly and this factor might also mediate the roll between learning a second language and executive functioning.
The second individual difference that will be looked at is quality of sleep. Jelicic et al. (2002) found that subjective complaints about sleep quality were negatively associated with cognitive functioning in older adults. So subjective sleep complaints could predict cognitive decline in elderly. In addition, Waller et al. (2016) found that self-reported poor sleep quality was related to changes in cognitive functioning in males in their late midlife, whereby a cognitively impaired group reported lower subjective sleep quality compared to a
cognitively improved group. Thereby, Ferrie et al. (2011) found that middle-aged adults that moved from a regular pattern of six to eight hours of sleep at a five year follow-up to she short and long ends of the sleep distribution had poorer cognitive function relative to those adults whose sleep distribution remained unchanged. Based on this research it seems that there is an important relation between cognitive functioning and quality of sleep, in which poor quality of sleep might result in worse cognitive functioning. Therefore, this factor potentially also contributes to the relation between learning a second language and executive functioning.
The third and last factor that will be looked at is physical activity. Colcombe & Kramer (2003) found, in a meta-analysis of eighteen studies, that physical activity had a positive effect on cognition, with the largest benefits occurring for executive-control processes. In a more recent study that used a physical activity intervention, Lautenschlager et al. (2008) found that adults who experienced subjective memory impairment had a modest improvement of
cognition after a physical activity program. In addition, Nagamatsu et al. (2013) found a positive influence of physical activity on cognitive functioning in elderly that were already experiencing cognitive decline. It seems that there is a link between physical activity and cognitive functioning. Therefore physical activity might also mediate the relationship between learning a second language and cognitive functioning.
In this study we will look at the influence of learning a second language on the executive functions switching, inhibition and updating, and the mediating role of intelligence, quality of sleep, and physical ability on this. We expect that a group elderly that learns Italian as a second language for three weeks, will all improve in their performance on a switching, an inhibition and an updating task and that the individual differences in intelligence, quality of sleep and physical activity will all have a positive mediating effect on this relationship.
Method
Subjects
For this study we selected 20 participants between 65 and 78 years of age. They were recruited using various methods: flyers were handed out in public locations, several community centres for the elderly were contacted and friends and acquaintances were asked for participation. All the participants lived in Amsterdam or surroundings. Participants were excluded if they were raised bilingual, if they spoke fluent Italian, if they had any (neuro)psychological disorder and if they were colour-blind.
Measures
Individual differences: Intelligence
To measure the intelligence of our participants we used the abstraction part of the Shipley Institute of Living Scale (Shipley) (Shipley, 1940). This
abstraction test is a good estimate of IQ scoring compared to the Wechsler Adult Intelligence Scale-Revised (WAIR-R) full-scale IQ (r = .64) (Paulson & Lin, 1970). It consists of 20 abstract thinking items, with a maximum score of 40 (every right answer is multiplied by two).
Individual difference: Quality of Sleep
To measure their quality of sleep (QoS) the participants filled in a logbook every day, between the two testing moments, asking them how they slept on a scale of one to ten. Their QoS score was then calculated by finding the mean sleep quality of the days they filled in the logbook, ranging between one and ten.
Individual difference: Physical Activity
To measure their physical activity (PA) the participants also filled in a logbook every day between the two testing moments, asking them if they had exercised that day and if this took more or less than 30 minutes. Their mean PA score was then calculated by finding the percentage of days they exercised more than 30 minutes of the days of their total days they filled in the logbook.
Language Course
For the language course we used an Italian language course, because this is a language not regularly taught in school and it has no negative connotations.
We gave participants the first four chapters of the course “Italian for Beginners”, which included learning simple words, sentences and grammar. Audio files were also included so the participant could practice with pronunciation. To test their Italian knowledge beforehand, we gave them a language test containing 32 items (words, small sentences and grammar) in which they either had to translate these from Dutch to Italian or from Italian to Dutch. Afterwards we tested their knowledge again using a similar language test with different items, to see how much their Italian improved over the three weeks time period. To exclude the possibility that one of the language tests was harder than the other, the order was randomized. We scored their test by awarding two points to a correct answer, one point if the answer that was almost right or written wrong and zero point if the answer was completely wrong.
Trail making task
For the switching task we used the trail making task (TMT). This task is also administered on a computer and consists of two parts, A and B. In part A the participant is asked to connect 25 consecutive numbers as fast as possible. In part B, the participant still has to connect the lines as fast as possible, but now alternating between consecutive numbers and letters (1-A-2-B-3-C… etc.). The two trails are equivalent except for an addition of a more complex cognitive task in part B, which requires switching. The TMT score is the difference between the reaction time for trail B minus trail A (Gaudino et al., 1995).
Simon task
a computer. Each trial started with a fixation cross in the centre of the screen and was followed by a blank interval. After this interval a red or a blue square
appeared on the screen, either on the left of right. Participants had to press either the left shift key, when they saw a blue square, or the rights shift key, when they saw a red square, regardless to where the square appeared (right or left). The task started with eight practice trials, followed by 72 experimental trials. These experimental trials were half congruent (coloured square appeared on the same side as the participant had to push the button, which does not require inhibition) and half incongruent (coloured square appeared on the opposite side as the participant had to push the button, which does require inhibition). The Simon effect was calculated by subtracting the reaction time in the incongruent trials minus the RT in the congruent trials, in which a lower score means less Simon effect (Bialystok et al., 2004).
Corsi Block Tapping Task
For our updating task we used the Corsi’s Block Tapping Task (Corsi). The Corsi was administered on a computer. In the first trial the participant saw two blocks light up one after the other and he or she had to repeat this sequence correctly. Two trials were given of every number of blocks and when at least one of them was repeated correctly, the next sequence would increase in length (so for the next trial there would be three blocks lighting up). When the participant did not repeat any of the two trials, of the same length, correctly, the test quit
automatically. The Corsi span is the number of blocks in a sequence a participant is able to correctly reproduce twice (Kessels et al., 2000).
Procedure
First, our participants filled in an online intake at home, to check for the exclusion criteria. After this, every participant who still met our criteria was invited for the first session (T0). During T0 our participants completed the abstraction part of the Shipley, to have an estimate of their IQ and next they completed the Simon task, the Corsi and the TMT. Then they completed the first Italian language test, to check for their Italian knowledge prior to the training. After all these tests were finished, they received an Italian language course. The language course could be used individually and at home, so our participants could work on this independently. They were supposed to follow this course for the next three weeks, half an hour a day, until their last testing moment. To measure the individual differences they filled in questions about QoS and PA in a logbook. After three weeks we had another testing session (T1) and all
participants completed the Simon Task, the Corsi and the TMT again. To check for their progression in Italian the participants also took an Italian language test again. Then they got an exit interview and they got informed about the purpose of our study.
Results
Due to various reasons, including not enough participation in the
language course and sickness, nine participants dropped out of the experimental condition. This left us with a final sample of thirteen participants to run our analysis on.
Language test
To check if our manipulation actually worked, we conducted a paired samples t-test. On average, people scored higher on the second language test (M = 26.15, SD = 9.56) than the first language test (M = 16.46, SD = 6.68). This difference was significant t(12) = -3.42, p = .005, so people got a significantly higher score on the second language test than on the first language test and our manipulation worked.
Trail Making Test
Based on a scatterplot, subject 7 and 10 showed such a big difference in their TMT score between T0 and T1 that it was likely that something went wrong when they took the test. Subject 7 had a TMT difference score of 112.91 seconds and subject 10 a TMT difference score of 146.67 seconds. Their scores also exceeded two or more standard deviations above the mean. For this reason it was decided to exclude their data as outliers. The final analysis of the TMT was run on 11 subjects.
First a paired-samples t-test was conducted to check if there was a significant difference between score on the TMT before and after the language course. The participants did have a smaller TMT after the language course (M = 24,84, SD = 16,59) than before (M = 27,80, SD = 19,15), which means that people got a lower difference between trail A and trail B on average after the language course. However no significant result was found between TMT at T0 and T1, t(10) = .58, p = .575. See also figure 1.
Figure 1. Average score on the TMT on the first and second measurement point. After this the individual differences mediating the relationship between acquire a second language and switching ability were looked at. The mean individual differences scores and their standard deviations for the TMT are shown in table 2.
Table 2
Means and standard deviations (in brackets) of the individual differences on the three different tasks. The Simon and Corsi have the same participants and therefore have the same scores on the individual differences, but the TMT has two outliers so consequently the individual differences scores are different.
Individual Differences
TMT (N=11)
Simon Effect & Corsi (N=13)
Shipley Score Average QoS Percentage PA 15.72(2.61) 7.27(0.99) 73.38(18.48) 15.31(4.24) 7.00(1.13) 70.72(18.59)
A multiple regression was conducted to see if these individual differences predicted the difference on the TMT score between T0 and T1.
Tests to see if the data met the assumption of collinearity indicated that multicollinearity was not a concern (Shipley scores, Tolerance = .73, VIF = 1.37; Quality of sleep scores, Tolerance = .80, VIF = 1.25; Physical activity percentages, Tolerance = .82, VIF = 1.22). The data also met the assumption of independent errors (Durbin-Watson value = 1.90). The histogram of standardised residuals indicated that the data contained approximately normally distributed errors, as did the normal P-P plot of standardised residuals. The scatterplot of
standardised predicted values showed that the data also met the assumption of homegeneity of variance and linearity. Lastly the data also met the assumption of non-zero variances (Shipley scores, Variance = 6.82; Quality of sleep scores, Variance = 0.98; Physical activity percentage, Variance = 341.53; TMT, Variance = 286.57).
Using the enter method it was found that the individual differences did not explain a significant amount of the variance in the difference on the TMT score (F(3,7) = 1.75, p = .24, R2 = .43 R2Adjusted = .18).
The analysis showed that Shipley score did not significantly predict a difference on the TMT score (β = -.44, t(7) = -.1.33, p = .23). In addition, QoS score did not significantly predict a difference on the TMT score (β = .07, t(7) = -.21, p = .84). Lastly PA percentage did not significantly predict a difference on the TMT score (β = .61, t(6) = 1.92, p = .10).
Simon Effect
Based on a scatterplot, subject 45 and 55 showed quite a big difference in their Simon effect between the first and second measurement. However they did not exceed two standard deviations above or below the mean, so it was decided to keep their data in the analysis. The final analysis was run on all 13 participants.
First a paired-samples t-test was conducted to check if there was a significant difference between the Simon effect before and after the language course. Participants did have less Simon effect after the language course (M = 25.08, SD = 91.88) than before (M = 74.28, SD = 105.90). However no significant result was found t(12) = 1.32, p = .21. See also figure 2.
Figure 2. Average Simon effect on the first and the second measurement point. After this the individual differences mediating the relationship between acquire a second language and inhibition ability were looked at. The mean individual differences scores and their standard deviations for the Simon task are shown in table 2.
A multiple regression was conducted to see if the individual differences predicted the difference of the Simon effect between T0 and T1.
Tests to see if the data met the assumption of collinearity indicated that multicollinearity was not a concern (Shipley scores, Tolerance = .87, VIF = 1.15; Quality of sleep scores, Tolerance = .89, VIF = 1.13; Physical activity percentages, Tolerance = .96, VIF = 1.04). The data also met the assumption of independent errors (Durbin-Watson value = 2.77). The histogram of standardised residuals indicated that the data contained approximately normally distributed errors, as did the normal P-P plot of standardised residuals. The scatterplot of
homegeneity of variance and linearity. Lastly the data also met the assumption of non-zero variances (Shipley scores, Variance = 10.88; Quality of sleep scores, Variance = 1.28; Physical activity percentage, Variance = 345.69; Simon, Variance = 17985.02).
Using the enter method it was found that the individual differences did not explain a significant amount of the variance in the difference of the Simon effect (F(3,9) = .37, p = .77, R2 = .11 R2Adjusted = -.19).
The analysis showed that score on the Shipley did not significantly predict a difference on the Simon effect score (β = .21, t(6) =.63, p = .55). In addition QoS score did not significantly predict a difference on the Simon effect (β = .20, t(6) = .60, p = .57). Lastly the PA percentage did not significantly predict a difference on the Simon effect (β = -.02, t(6) = -.06, p = .96).
Corsi Block Tapping Test
Based on a scatterplot, there were no outliers for the Corsi. So the final analysis was run on all 13 subjects.
First a paired-samples t-test was conducted to check if there was a significant difference between Corsi span before and after the language course. The participants did have a slightly higher span after the language course (M = 5.23, SD = .93) than before (M = 5.08, SD = .93). However no significant result was found, t(12) = -.40, p = .700. See also figure 3.
Figure 3. Average Corsi Span on the first and the second measurement point. After this the individual differences mediating the relationship between acquire a second language and updating ability were looked at. The mean individual differences scores and their standard deviations for the Corsi are shown in table 2.
A multiple regression was conducted to see if the individual differences predicted the difference on the Corsi span between the first and the second test moment.
Tests to see if the data met the assumption of collinearity indicated that multicollinearity was not a concern (Shipley scores, Tolerance = .87, VIF = 1.15; Quality of sleep scores, Tolerance = .89, VIF = 1.13; Physical activity percentages, Tolerance = .96, VIF = 1.04). The data also met the assumption of independent errors (Durbin-Watson value = 1.68). The histogram of standardised residuals indicated that the data contained approximately normally distributed errors, as
did the normal P-P plot of standardised residuals. The scatterplot of
standardised predicted values showed that the data also met the assumption of homegeneity of variance and linearity. Lastly the data also met the assumption of non-zero variances (Shipley scores, Variance = 10.90; Quality of sleep scores, Variance = 1.28; Physical activity percentage, Variance = 345.69; TMT, Variance = 1.97).
Next a multiple regression was conducted to see if the individual differences predicted the difference of the Corsi span between T0 and T1.
Using the enter method it was found that the individual differences did not explain a significant amount of the variance in the difference of the span (F(3, 9) = .69, p = .58, R2 = .19 R2Adjusted = -.08).
The analysis showed that score on the Shipley did not significantly predict a difference on the Corsi span (β = -.35, t(6) = -1.10, p = .30). In addition, QoS score did not significantly predict a difference on the Corsi span (β = .16, t(6) = .50, p = .63). Lastly, the PA percentage during the testing period did also not significantly predict a difference on the Corsi span (β = .32, t(6) = 1.04, p = .33).
Discussion
The aim of our study was to investigate the mediating role of individual differences on the acquiring of a second language on executive functioning. However the language course did not have a significant effect on executive functioning; the inhibition, switching and updating ability of our participants did not increase after acquiring a second language. Furthermore, the three individual differences did not have a significant mediating effect on this relationship.
However knowledge of Italian did significantly improve between the first and the second measurement.
The significant improvement of Italian knowledge made it seem like the manipulation worked. However even though there was a significant difference between knowledge of Italian before and after the course, the participants still scored on average only half of the points attainable. Consequently, they did probably not acquire Italian as a fluent second language, and therefore it cannot be concluded that have become bilingual and obtained the executive functioning advantages found in bilinguals. According to Jackson & Kaplan (1999) it takes approximately 23-24 weeks (or 575-600 class hours) before a second language is learned. So a time frame of three weeks, half an hour a day, does presumably not result in acquiring a second language to the extend that the bilingual advantages will be achieved. As a proposal for a future research it would be useful to give the participants a longer time frame in which they can acquire the second language.
Yet, individual differences mediating the relationship between learning a second language and executive functioning could still be expected. Figure 4 to 6 show that there was a lot of variability within the individual improvement on the three executive functions: some participants did improve their executive
Figure 4. Individual scores between T0 and T1 on the TMT, a lower score implying a better switching ability
Figure 5. Individual scores between T0 and T1 on the Simon, a lower score implying a better inhibition ability
Figure 6. Individual scores between T0 and T1 on the Corsi, a higher score implying better updating ability
However no effects were found of any of the individual differences on the relationship between acquiring a second language and the three different
executive functions. Reasons why there might not have been an effect of these differences are offered below.
A potential reason why there was no effect found of intelligence could be that our participants only completed the abstraction part of the Shipley and not a full intelligence test. As a result, their intelligence was not measured completely. The Shipley abstraction is a measurement of reasoning ability, so this probably does not measure all the aspects of the general intelligence (Salthouse et al., 1999). Even though the Shipley abstraction scores do have quite a high correlation with WAIS full scale IQ (r = .64) (Paulson & Lin, 1970), this result might be dated. As Flynn (1984, 1987, 1999; retrieved from Must et al. 2003) pointed out, the IQ of the population of several countries has increased by about
three points a decade for the last 60 years, which is now also known as the Flynn effect. Because of this it might be that the Shipley, which was created back in 1940, does not measure peoples intelligence nowadays accurately. A suggestion for future research would be to let the participants complete a more accurate and complete intelligence test to be able to have a better estimate of a persons intelligence.
Potential reasons why QoS did not account for the differences found between acquiring a second language and executive functioning as well, could either be that QoS simply does not have a significant effect on the improvement of cognitive functions after a language course or it could be that
operationalization of QoS in this study did not measure the real QoS of our participants. Based on earlier research (Jelicic et al., 2002; Waller et al., 2016; Ferrie et al., 2011) this second reason would be more plausible. To measure QoS, the participants rated their QoS on a scale of one to 10, which is not a valid scale for QoS. As a suggestion for future research QoS should be measured with a more valid scale, for instance the subscale Sleep Problems of the Symptoms Checklist – 90 (used in Jelicic et al., 2002) to assess peoples sleep complaints, or the
Pittsburgh Sleep Quality Index, measuring sleep quality and disturbances (used in Waller et al., 2016).
Lastly, no effects of PA were found mediating the relationship between learning a second language and executive functioning. However based on previous research, this was expected (Colcombe & Kramer, 2003; Nagamatsu et al., 2013; Lautenschlager et al.,2008). A potential reason for this could be that PA was also measured not using a valid scale, but by asking our participants if they exercised or not and if it was for longer or shorter than 30 minutes. In addition,
no good definition of PA was given. Due to this, interpretation of this concept might have differed between subjects. Therefore a potential point of
improvement for the next research could be to more concretely define PA, or even work with a point-system where different physical activities get more or less points, depending on the amount of energy expenditure they require (for example walking half an hour will get one point and jogging will get two).
Another way to improve the physical activity measurement would be to have an experimental exercise group and a control group, within the second-language learning group. This way the amount of exercise a group will participate in can actually be controlled for. For instance, the experimental group could get a workout program for at least half an hour every day, while the control group would not. This way better statements could be made about the influence of physical activity on the relationship between learning a second language and executive functioning.
Thus, the operationalization of these three individual differences could be the reason they did not account for any of the differences found in the
relationship of acquiring a second language and executive functioning. Future studies could try to unravel the reasons behind this variation and create a more complete picture why some people improve a lot in their executive functioning due to learning a second langue and others do not. However, it is important to note that this study did not use a control group, so it is unclear if the
improvement of some individuals on their executive functioning is due to the acquiring of a second language, or due to learn-effects. The use of a control group could also be a recommendation for future studies.
Another major problem concerning our research, not only for finding an effect of acquiring a second language on executive functioning but also for the individual differences mediating this role, is the sample size. Using a sample size calculator (Soper, 2006), with a power of .8, an effect size of .15 and a probability level of .05, the minimum required sample size is 76. Due to dropout, exclusion and shortage of time, a sample of only 13 participants remained, so the final analysis of our study was run on less than one fourth of the required
participants. Consequently, even if there was any effect of the three-week language course on executive functioning and of the mediating role of the individual differences, due to the small sample size, no effects were actually found. In addition, the results found cannot be generalized to the whole
population, because such a small sample can never be a true representation of the entire population of elderly.
Because our research has so many pitfalls, it would be accurate to
consider it a pilot study. None of our hypotheses have been confirmed, but due to the small sample size and the short time frame to acquire the second language, being able to confirm the hypotheses was not to be expected. In addition, our individual differences did not explain the variability found between the executive functioning ability of our participants between the two testing moment, but this could be due to poor operationalization. Nonetheless, the assumption that acquiring a second language could have a positive effect on executive functioning has quite some probability based on earlier research (Gold et al., 2013; Bialystok et al., 2004; Luo et al., 2013; Bialystok & Craik, 2010; Soveri et al., 2011). Therefore, it would not be right to discard the whole theory based on the current study. Using a bigger sample size and giving the
participants a longer time period to acquire the language, could still improve executive functioning of the whole group. In addition, a better operationalization of the individual differences might be able to explain the differences in
improvement on the executive functions. Due to the major relevance it has to society, it would still be really useful to further investigate this theory. Not only is the elderly population increasing in age (Wilson et al., 2002), they are also expected to almost double in number; from 2.7 million in 2012 to 4.7 million in 2041 (Giesbers et al., 2014). This means that we are facing an enormous growth of the elderly population, that, in addition, has a higher life expectation. Since cognitive decline is not only universal, but also correlated with lower quality of life (Mancke et al., 2006), finding ways to lessen this decline are of great
importance. Since bilingualism has so many advantages on executive functioning, which is also one of the first cognitive functions to decline at older age (Treitz et al., 2007), finding a way to transfer these advantages to monolinguals could have a major effect on the elderly population. By using this study as a pilot, and
applying the recommendations that were provided, future studies could find a way to improve the cognitive decline in elderly and provide them with a long and cognitive functional life.
References
Ball, K., Berch, D. B., Helmers, K. F., Jobe, J. B., Leveck, M. D., Marsiske, M., .Morris, J. N., Rebok, G. W., Smith, D. M., Tennstedt, S. L., Unverzagt, F. W., & Willis, S. L. (2002). Effects of Cognitive Training Interventions With Older Adults: a randomized controlled trial. Jama, 288(18), 2271-2281.
Bialystok, E., & Craik, F. I. M. (2010). Cognitive and linguistic processing in the bilingual mind. Current directions in psychological science, 19(1), 19-23.
Bialystok, E., Craik, F. I. M., Klein, R., & Viswanathan, M. ( 2004). Bilingualism, aging, and cognitive control: evidence from the simon task. Psychology and aging, 19(2), 290-303.
Bolla, K., Brown, K., Eldreth, D., Tate, K., & Cadet, J. (2002). Dose-related neurocognitive effects of marijuana use. Neurology, 59(9), 1337-1343.
Buitenweg, J. I. V., Murre, J. M. J., & Ridderinkhof, K. R. (2012). Brain training in progress: a review of trainability in healthy seniors. Frontiers in Human Neuroscience, 6(183), 1-11.
Chapman, S. B., Aslan, S., Spence, J. S., Hart, J. J., Bartz, E. K., Didehbani, N., .Keebler, M. W., Gardner, C. M., Strain, J. F., DeFina, L. F., & Lu, H. (2013). Neural Mechanisms of Brain Plasticity with Complex Cognitive Training in Healthy Seniors. Cerebral Cortex, 25(2), 396-405.
Cheng, Y., Wu, W., Feng, W., Wang, J., Chen, Y., Shen, Y., Li, Q., Zhang, X., & Li, C. (2012). The effects of multi-domain versus single-domain cognitive
training in non-demented older people: A randomized controlled trial. BMC Medicine BMC Medicine, 10(1), 1-13.
Colcombe, S., & Kramer, A. F. (2003). Fitness effects on the cognitive function of older adults: a meta-analytic study. Psychological science, 14(2), 125-130.
Ferrie, J. E., Shipley, M. J., Akbaraly, T. N., Marmot, M. G., Kivimäki, M., & Singh-Manoux, A. (2011). Change in sleep duration and cognitive function: findings from the Whitehall II study. SLEEP, 34(5), 565-573.
Gaudino, E. A., Geisier, M. W., & Squires, N. K. (1995). Construct validity in the trail making test: what makes part B harder? Journal of Clinical
Experimental Neuropsychology, 17(4), 529-535.
Giesbers, H., Verweij, A., & Beer, de, J. (2014). Vergrijzing: Wat zijn de
belangrijkste verwachtingen voor de toekomst? Volksgezondheid Toekomst Verkenning, Nationaal Kompas Volksgezondheid. Bilthoven, RIVM, Nationaal Kompas
Gold, B. T., Kim, C., Johnson, N. F., Kryscio, R. J., & Smith, C. D. (2013). Lifelong Bilingualism Maintains Neural Efficiency for Cognitive Control in Aging. The Journal of Neuroscience, 33(2), 387-396
Jackson, F. H., & Kaplan, M. A. (1999). Lessons learned from fifty years of theory and practise in government language teaching. Washington DC, Georgetown University Round Table on Languages and Linguistics, 71-87.
Jelicic, M., Bosma, H., Ponds, R. W. H. M., Boxtel, Van, M. P. J., Houx, P. J., & Jolles, J. (2002). Subjective sleep problems in later life as predictors of cognitive decline. Report from the Maastricht Ageing Study (MAAS). International Journal of Geriatric Psychiatry, 17, 73-77.
Kessels, R. P. C., Zandvoort, Van, M. J. E., Postma, A., Kappelle, L. J., & Haan, De, E. H. F. (2000). The Corsi-Block Tapping Task: standardization and normative data. Applied neuropsychology, 7(4), 252-258.
Kramer, R., & Mota, M. B. (2015). Effects of bilingualism on inhibitory control and working memory: a study with early and late bilinguals. Gragoatá, 38, 309-331.
Lautenschlager, N. T., Cox, K. L., Flicker, L., Foster, J. K., Bockxmeer, Van, F. M., Xiao, J., Greenop, K. R., Almeida, O. P. (2008). Effect of physical activity on cognitive function in older adults at risk of Alzheimer disease. JAMA, 300(9), 1027-1038.
Luo, L., Craik, F. I. M., Moreno, S., & Bialystok, E. (2013). Bilingualism interacts with domain in working memory task: evidence from aging. Psychology and aging, 28(1), 28-34.
Lustig, C., Shah, P., Seidler, R., & Reuter-Lorenz, P. A. (2009). Aging, training, and the brain: a review and future directions. Neuropsychol. Rev., 19(4), 504-522.
Mahncke, H. W., Connor, B. B., Appelman, J., Ahsanuddin, O. N., Hardy, J. L., Wood, R. A., Joyce, N. M., Boniske, T., Atkins, S. M., & Merzenich, M. M. (2006). Memory enhancement in healthy older adults using a brain plasticity-based training program; a randomized, controlled study. PNAS, 103(33), 12523-12528.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, A. (2000). The unity and diversity of executive functions and their contributions to complex “frontal lobe” tasks: a latent variable analysis. Cognitive
Psychology, 41, 49-100.
Murayama, N., Iseki, E., Tagaya, H., Ota, K., Kasanuki, K., Fujishiro, H., Arai, H., & Sato, K. (2013). Intelligence or years of education: which is better
correlated with memory function in normal elderly Japanse subjects? Pyschogeriatrics, 13, 9-16.
Must, O., Must, A., & Raudik, V. (2003). The secular rise in IQs: In Estonia, the Flynn effect is not a Jensen effect. Intelligence, 31, 461-471.
Nagamatsu, L. S., Chan, A., David, J. C., Beattie, B. L., Graf, P., Voss, M. W., Sharma, D., & Liu-Ambrose, T. (2013). Physical activity improves verbal and spatial
memort in older adults with probable mild cognitive impairment: a 6 month randomized controlled trial. Journal of Aging Research, 2013, 1-10.
Paulson, M. J., & Lin, T. (1970). Predicting WAIS IQ from Shipley-Hartford scores. Journal of Clinical Pyschology, 26(4), 453-461.
Salthouse, T. A. (1996). Constraints on theories of cognitive aging. Psychonomic Bulletin & Review, 3(3), 287-299.
Schlee, W., Leirer, V., Kolassa, I., Weisz, N., & Elbert, T. (2012). Age-related changes in neural functional connectivity and its behavioural relevance. BMC Neuroscience, 13(16), 1-11
Shipley, W. C. (1940). A self-administering scale for measuring intellectual impairment and deterioration. The Journal of Psychology, 9(2), 371-377.
Soper, D. (2006). Calculator: a-priori sample size for multiple regression. Received from: http://www.danielsoper.com/
Soveri, A., Rodriguez-Fornells, A., & Laine, M. (2011). Is there a relationship between language switching and executive functions in bilingualism? Introducing a within-group analysis approach. Frontiers in psychology, 2(183), 1-8.
Stern, Y. (2002). What is cognitive reserve? Theory and research application of the reserve concept. Journal of the International Neuropsychological Society, 8, 448-460.
Treitz, F. H., Heyder, K., & Daum, I. (2007). Differential course of executive
control changes during normal aging. Aging, neuropsychology and cognition, 14, 370-393.
Waller, K. L., Mortensen, E. L., Avlund, K., Osler, M., Fagerlund, B., Lauritzen, M., & Jennum, P. (2016). Subjective sleep quality and daytime sleepiness in late midlife and their association with age-related changes in cognition. Sleep Medicine, 17, 165-174.
Wilson, R. S., Beckett, L. A., Barnes, L. L., Schneider, J. A., Bach, J., Evans, D. A., & Bennet, D. A. (2002). Individual differences in rates of change in cognitive abilities of older persons. Psychology and aging, 17(2), 179-193.
