Introduction:

Bilingualism is defined as the usage and proficiency in at least two languages by an individual, which may change depending on the opportunities to use the languages and exposure to other users of the languages (ASHA, 2004). Haugen (1953) asserted that bilingualism starts when an individual communicating in one language can equally communicate in another language using meaningful utterances. Innumerable definitions have led to the emergence of a number of classification systems to describe the various types of bilingualism.

Based on the age of exposure to the two languages, bilinguals are classified as early and late bilinguals (Beardsmore, 1986), depending on how two or more linguistic codes are organized and stored by individuals, bilinguals are grouped as compound, coordinate and subordinate bilinguals (Weinreich, 1957) and based on the proficiency and the competency in both the languages, bilingualism has been classified into balanced and dominant types (Peal & Lambert, 1962). When L1 competence is equal to L2 competence, the condition is called balanced type and in the dominant type the L1 competence is greater than or less than the L2 competence.The competency may include a variety of skills, and since proficiency varies across the two languages at all linguistic levels for a bilingual, proficiency has to be assessed in a variety of domains such as understanding, speaking, reading and writing in both the languages. Many measures have been proposed to measure proficiency levels. McNamara (1967) grouped the measures of bilingual ability which assesses proficiency into mainly four categories, which include fluency tests, rating scales, dominance tests and flexibility tests. Among the many rating scales available Language Experience and Proficiency Questionnaire developed by Flege, 1999 and revised by Marian, Blumenfeld & Kaushanskaya, 2007 is a valid and reliable questionnaire for assessment and quantification of a bilingual’s proficiency. It has also been adapted to the Indian context by Ramya & Goswami (2009)

Bilingualism has been found to have a positive effect in the efficiency of the executive functioning (EF) system. “Executive functions” (EFs), is a broad term which includes many cognitive processes and behavioral competencies (Chan, Shum & Chen, 2008). Cognitive flexibility, response inhibition or resistance to interference, verbal reasoning, problem-solving, planning, resistance, feedback utilization, multitasking sequencing, attention skills, and the ability to deal with novelty. Executive functioning in bilinguals refers to higher processes that involve conscious control of thought and action (Zelazo & Muller, 2010).Bilinguals have the advantage of being able to constantly shift between two languages and paying attention in selecting the required word from a set of distracting stimulus which boosts their cognitive development. Bilinguals took lesser time as compared to the monolinguals in the switching tasks, indicative of the fact that better experience in can lead to better efficiency in the ability to shift flexibly between mental sets. The advantages of bilingualism in executive function are not only applicable to response inhibition tasks but also to flexibility in cognitive switching. Bilingual advantages in executive function which is an emerging area in the field of cognitive development can be assessed by tapping various domains of response inhibition, cognitive flexibility, attention and working memory using tasks which could be verbal or nonverbal. (Bialystok, 2001)

Response Inhibition or interference refers to the ability to block the extra information in order to focus on a particular task.  There are many tasks to measure inhibition in bilinguals. One such is the Stroop task where there would be asterisks appearing in colored ink along with color words appearing in both congruent and incongruent ink. The difference between the speed of responding to all the asterisks and the speed for naming the incongruent stimuli will be considered as the Stroop effect. Simon’s task and Flankers task can also be used to test the same.

Cognitive flexibility or set shifting is the ability of an individual to shift quickly between different response sets (Anderson, 2002). Prior & MacWhinney in 2010 claimed that this switching requires the selection of a situation-appropriate language between two active language systems. There are a series of nonverbal tasks also such as the local-global task where a global larger figure composed of local smaller figures would be presented. This task comprises of both congruent and incongruent trials and the difference between the speed of responding for a congruent trial from the time that of an incongruent trial will be considered as the shifting effect. 

Attention refers to how an individual can actively process specific information in the environment. Attention in bilinguals can be measured using Navon’s figures where a bigger recognisable shape, such as a letter is presented. The larger letter will comprise of smaller different shapes and the subjects have to pay attention and perceive the global as well as the local features of the items presented.

Different linguistic and non linguistic tasks have been employed to test executive functions. Non linguistic tasks includes Flankers task, Stroop object Task , Attentional network task (ANT) , Simons task while linguistic tasks include priming experiments which tap for listening comprehension, dichotic listening tasks and Stroop task etc.

Very few studies on executive functions in high proficient and low proficient bilinguals have been carried out and hence the current study was conducted to explore the effects of bilingualism on the two groups of bilinguals in the various domains of executive functions namely attention, response inhibition and cognitive flexibility.

Need for the study:

Executive functions are considered as a critical component of cognitive and social development in bilinguals. The research done in the field of bilingualism has many clinical implications for understanding the development of executive function and also the various practical issues with respect to second-language instruction in homes and schools. India is a country having bilinguals of varied proficiency. Executive functions vary as a function of degree of bilingual proficiency. Till date a very few attempts have been made to assess the effects of bilingualism on executive functioning in the Indian context.  There are many domains of executive functioning, however this study is aimed at studying the effects of bilingualism specifically on the domains of cognitive flexibility, response inhibition and the ability to sustain attention as these tasks are sensitive to vary with respect to high and low proficient bilinguals by using tasks of varying complexity for each of the domains.

Method:

Executive functions are considered as a critical component of cognitive and social development in bilinguals and it has been found that executive functions vary as a function of bilingual proficiency. There are many domains of executive functioning, however this study is aimed at studying the effects of bilingualism specifically on the domains of cognitive flexibility, response inhibition and the ability to sustain attention as these tasks are sensitive to vary with respect to high and low proficient bilinguals.

Objectives of the study were as follows:

1. To study the attention in high and low proficient bilinguals.
2. To study the response inhibition in high and low proficient bilinguals.
3. To study the task switching ability in high and low proficient bilinguals.

Participants:

Participants in the age range of 18-25 years were considered for the study. They were divided into 2 groups having 20 participants each which include equal number of both males and females.

Participant selection criteria:

Inclusionary criteria:

• All the participants were successive bilinguals having Kannada as L1 and English as their L2.
• All of the participants had exposure to L2 (English) right from their childhood with a minimum of 10 years of exposure to the language.
• The participants were tested through informal screening for normal vision.

Administration of LEAP-Q: Classification of Bilinguals:

Language Experience And Proficiency Questionnaire - LEAP Q is a tool to assess bilingual proficiency which was developed by Flege, 1999 and revised by Marian, Blumenfeld & Kaushanskaya, 2007. It has also been adapted to the Indian context by Ramya & Goswami (2009). The questionnaire contains 18 questions eliciting details about language acquisition and usage which was used to determine bilingual proficiency. Question 10 of the questionnaire required the participants to rate their proficiency on four domains: understanding, speaking, reading and writing using a four point rating scale (where, 1-Zero Proficiency, 2-Low, 3-Good and 4-Perfect Proficiency).  Based on LEAP-Q, Hayward (2013) claimed that if a bilingual has a score of 3 or 4 on the speaking domain of L2 they can be classified as high proficient bilinguals whereas Hickey (2010) claimed that a bilingual should receive a score of 4 in the understanding domain and a minimum score of 3 on all the other domains (speaking, reading and writing) in order to be classified as a high proficient bilingual. Based on the ratings done by the participants they were classified as high proficient or low proficient bilinguals by adopting Hayward’s criterion and were put into group 1 and group 2 respectively.

Tasks:

There were 2 tasks of varying complexity in each of the three domains tapping for attention (Tasks 1 and 2), response inhibition (Tasks 3 and 4) and cognitive flexibility (Task 5 and 6).

In addition to these tasks, alternating fluency task was also used and it is assumed to tap for all the three processes (attention, response inhibition and cognitive flexibility).

Instrumentation:

The stimulus for all the tasks was presented in visual mode on a 15.6 inch laptop through the DMDX software.

The participants were made to sit at a distance of 50cm from the laptop screen and the testing was done in a silent room. Instructions varied with respect to each task and are described under each task.             

Procedure:

Task 1 for Attention: Shapes Task:

Stimulus:  Different shapes were used which was chosen indiscriminately.

Task description: Two shapes were displayed on the computer screen. If both the shapes are the same the condition would be a congruent trial and if both the shapes appearing are different then the condition would be an incongruent trial.

Procedure: The participant was instructed to press 1 for congruent trial and 0 for incongruent trial.

Analysis: Reaction time and accuracy was measured for this task

Example: Both the shapes appearing were circles - congruent trial.

   

                The two shapes appearing was a circle and triangle- incongruent trial

Task 2 for Attention: Letter monitoring task:

Stimulus: The letter /p/ was chosen as the target and the words were chosen from a study done by Yashaswini & Abhishek (2015)

Task description: The stimulus used was English letters a pre set phoneme (/p/ for example was chosen. The task of the participants was to verify if the presented words have the target letter as its constituent regardless of the loci. The participant had to press the button 1 in the key board if the pre-set phoneme is present and 0 if the word does not contain the preset phoneme. The order of presentation of the stimulus was random.

Procedure: The participant was asked to press 1 if the letter and its constituents are same (congruent trial) and 0 if the letter and its constituents are different (incongruent trial). 

Analysis: Reaction time and accuracy was assessed for the task.

Example

Pet, cup etc-The participant had to press 1 as the pre set letter (target) is present in these words.

Goat, Mat etc- The participant had to press 0 as the pre set letter (target) is absent in these words

Task 3 for Response inhibition: Flankers Task:

Stimulus: The stimulus was arrow marks which were chosen indiscriminately.

Task description:  The central target was aligned towards the right or left side. The participant was required to look for the direction of the central arrow and press the keys based on the alignment of the stimulus. 

Procedure: The participant was asked to press the right key for congruent flankers and left key for incongruent flankers.

Analysis: The reaction time and accuracy was calculated here.

Example:

 

Task 4 for Response inhibition: Picture Naming Task:

Stimulus: The pictures for the task were chosen indiscriminately and the pictures was accompanied with a blue or red dot.

Task description: A picture appeared on the computer screen with the presence of a red or a blue dot. Depending on the color of the dot the participant had to name the picture. The stimulus was presented randomly and reaction time and accuracy was calculated.

Procedure: The participant had to name the item in Kannada if the picture was accompanied by a blue dot and in English if the picture was accompanied by a red dot.

Analysis: Reaction time and accuracy was measured for this task.

Example:  

Task 5 for Cognitive flexibility: Stroop Object Task

Stimulus: The stimulus used was taken from a study done by Pavithra & Prema (2015). The stimulus was pictures of certain items in either congruent or incongruent color ink.

Task description: Randomized presentation of the pictures was done with a few items shown in different color ink and few in congruent ink.

Procedure:  The participants were instructed to press 1 if it is a congruent condition and 0 for incongruent condition.

Analysis: Reaction time and accuracy for all of the trials was calculated.

Example: A picture of an apple was shown in red as well as blue

 

 

Task 6 for Cognitive flexibility: Letter Number Task

Stimulus: The stimulus included numbers in English and letters in both Kannada and English.

Task description: In this task the number was presented in English and was kept constant throughout. The letters presented alongside the number was in either Kannada or English. The stimulus included both congruent and incongruent trials and was presented in random order.

Procedure:  The participant was asked to press 1 for congruent condition and 0 for incongruent condition.

Analysis: Reaction time and accuracy was measured.

Example: 1 E (Congruent trial) and 6   (Incongruent trial)

Task 7 for Alternating fluency task:

Stimulus: 3 Lexical categories was considered (Animals, fruits, vegetables)

Task description: Under each lexical category, the person had to name a lexical item in Kannada and name one lexical item in English. The participant was instructed to adhere to the lexical category and name different lexical items in both the language (not translation equivalents)

Procedure: The number of correct lexical items named in L1 (Kannada) and L2 (English) mentioned under each lexical category was taken into consideration

Analysis: Quantitative analysis was carried out

Example: If the prescribed lexical category is animals, correct response would like /kudure/ and /lion/ which are distinctive responses and incorrect response would be like /ka:ge/ and /lion/ (as they belong to distinct lexical category or /kudure/ and /horse/ (as they are translation equivalents)

TASK	STIMULUS	DOMAIN ASSESSED
Task 1	Different shapes	Attention , Perception
Task 2	Letter monitoring (English letters with and without pre set target letter)	Attention, Perception
Task 3	Arrows	Response inhibition
Task 4	Pictures accompanied by a color dot	Response inhibition
Task 5	Pictures	Cognitive flexibility , Selective attention
Task 6	English numbers and letters in Kannada and English	Cognitive flexibility
Task 7	Alternating fluency	Attention, Response Inhibition and Cognitive flexibility

Table 1: Description of tasks

Scoring and Analysis:

Statistical analysis for calculating the mean reaction time and accuracy was done using the software Statistical Package for Social Sciences (SPSS) version 16.0.Shapiro Willis test for normality was administered for the high proficient and the low proficient group for all the parameters. In addition a few non parametric tests were also administered. Mann Whitney U test was administered to see the significant differences between males and females and also for checking the difference between high proficient and low proficient groups. Wilcoxon’s Signed Rank test was administered for comparison between Task 1 and Task 2 for the three domains irrespective of proficiency .Friedman test was administered for comparisons across the alternating fluency tasks.

Results and Discussion:

The primary measure the effects of bilingualism on executive aim of the study was to functions specifically on the domains of cognitive flexibility, response inhibition and the ability to sustain attention as these tasks are sensitive to vary with respect to high and low proficient bilinguals.

The results of the present study are discussed under the following headings:

1. Comparison of reaction time and accuracy between the high proficient and low proficient group in the domains of

1. Attention
2. Response Inhibition
3. Cognitive flexibility (between groups comparison)

 

2. Comparison between Task 1 and Task 2 for the three domains for both high proficient and low proficient group across mean reaction time and accuracy (within group comparison).

 

4. Comparisons across the alternating fluency tasks between the high proficient and low proficient group (between groups comparison)

Objective 1:

1. Comparison of reaction time and accuracy between the high proficient and low proficient group on the domains of

1. Attention

              The overall mean, standard deviation (SD) and median were calculated for the

performance of Group I (High proficient) and Group II (Low proficient) across the two tasks for both mean reaction time and accuracy.

Table 2 shows the performance of both the groups across the two tasks.

Tasks                                   High proficient                                                  Low Proficient                                  Mean         SD          Median                         Mean        SD        Median Task 1                    651.56       51.18          638.62                       745.86        74.43      717.61 Task 2                    712.96       76.22          692.28                       704.19        82.42       681.93

Table 2: Mean, SD and Median for  mean reaction time for Group I (High proficient) and Group II (Low proficient) across the two tasks of attention.

As depicted in Table 2 the mean reaction time taken by the high proficient group (Mean = 651.56, Median = 638.62) was lesser than the low proficient group (Mean = 745.86 , Median = 717.61) on Task 1 of attention. However for Task 2 of attention the mean reaction time of the high proficient group was found to be slightly higher (Mean = 712.96 , Median = 692.28)  than the low proficient group (Mean = 704.19 , Median=681.93). In reaction time studies, lesser the time taken to perform the task, the better is the performance. SD was more for low proficient group compared to high proficient group.

Tasks                                   High proficient                                                  Low Proficient                                  Mean         SD          Median                         Mean        SD        Median Task 1                    96.44          2.01           96.60                       93.96            4.26         94.95 Task 2                    92.96          3.40           93.30                       86.63            3.42         86.60

Table 3: Mean, SD and Median for accuracy for Group I (High proficient) and Group II (Low proficient) across the two tasks of attention.

In addition to mean reaction time, accuracy of responses was also computed. Accuracy scores for the high proficient group (Mean = 96.44, Median = 96.60) was greater compared to the low proficient group (Mean = 93.96, Median = 94.95) on Task 1. Scores obtained on Task 2 also followed a similar trend, (Mean = 92.9, Median =93.30) for high proficient group and (Mean = 86.63,  Median = 86.60 for low proficient group.

In order to verify if there was any significant difference between high and low proficient groups on Task 1 of attention, Mann Whitney U Test was carried out. While measuring the reaction time the |Z| score obtained was 3.841 and the corresponding p value showed no significant difference. On measuring accuracy the |Z| score obtained was 1.991 and the corresponding p value showed significant difference. On Task 2 of attention the |Z| score obtained for mean reaction time was 0.541 and for accuracy 4.451 and the corresponding p values for both the measures showed significant differences.

To summate the high proficient group outperformed the low proficient group in terms of accuracy of responses as there was a statistically significant difference (p<0.05) for both the tasks. In regard to reaction time, statistically significant difference was seen for only task 1 of attention and for task 2 no significant difference (p>0.05) was seen.

Comparison of reaction time and accuracy between the high proficient and low proficient group on the domains of

2. Response Inhibition

Tasks                                   High proficient                                     Low Proficient                                  Mean          SD          Median                    Mean           SD              Median Task 1                    743.22         119.04      738.03                   719.60           89.78            711.99 Task 2                    1089.97        95.82      1076.23                 1120.75        112.31          1123.22

Table 4: Mean, SD and Median for mean reaction time for Group I (High proficient) and Group II (Low proficient) across the two tasks of response inhibition.

As shown in Table 4 the high proficient group (Mean =743.22 , Median = 738.03) took more time than the low proficient group (Mean = 719.60 , Median = 711.99) on Task 3 of response inhibition.  For Task 4 of response inhibition the mean reaction time of the high proficient group Mean= 1089.97, Median = 1076.23) was lesser than the low proficient group (Mean = 1120.75 ,Median=1123.22).

__________________________________________________________________________________________

                                 Mean         SD          Median                   Mean           SD        Median

Task 1                    92.29          4.06          93.30                       91.46           3.97        93.30

Task 2                    92.31          3.75          93.30                       87.30           5.25        86.60

__________________________________________________________________________________________

Table 5: Mean, SD and Median for accuracy for Group I (High proficient) and Group II (Low proficient) across the two tasks of response inhibition.

As depicted above, the accuracy of responses for both the groups was noted.  The high proficient group (Mean= 92.29) had better accuracy of responses than the low proficient group (Mean = 91.46) on both Task 3 and Task 4 Mean = 92.31 for high proficient group and Mean = 87.30 for low proficient group on the domain of response inhibition. The median values also followed the same direction.

In an attempt to verify if there was a significant difference between high and low proficient groups on Task 3 of response inhibition, Mann Whitney U Test was carried out. While measuring the reaction time the |Z| score obtained was 0.568 and the corresponding p value showed no significant difference. While measuring accuracy the |Z| score obtained was 0.730 and the corresponding p value also showed no significant difference. On Task 4 of response inhibition the |Z| score obtained for mean reaction time was 0.649 and the p value showed no significant difference, however for accuracy the |Z| value obtained was 3.218 and the corresponding p value showed significant difference.

In summary statistically significant difference was seen only on accuracy scores of Task 4 (p<0.05). Mean reaction time for high proficient and low proficient group was not significant statistically for both task 3 and task 4. Statistically significant difference was not seen between the high and low proficient groups in regard to accuracy scores for task 3.  

Comparison of reaction time and accuracy between the high proficient and low proficient group on the domains of

3. Cognitive Flexibility

________________________________________________________________________________________

                                 Mean         SD          Median                         Mean        SD        Median

Task 5                    730.81       89.06          714.86                       719.42        79.79      703.20

Task 6                    730.39       119.32        703.94                       730.17        73.30      723.55

__________________________________________________________________________________________

Table 6:Mean, SD and Median for mean reaction time for Group I (High proficient) and Group II (Low proficient) across the two tasks of cognitive flexibility

Table 6 shows the mean reaction time taken by the high proficient group (Mean = 730.81 , Median = 714.86) which was found to be greater than the low proficient group (Mean = 719.42  , Median = 703.20) on Task 5 of cognitive flexibility. For Task 6 of cognitive flexibility the mean reaction time of the high proficient group was in par (Mean = 730.39 , Median = 703.94) with the low proficient group (Mean= 730.17 , Median= 723.55).

__________________________________________________________________________________________

                                 Mean         SD          Median                         Mean         SD         Median

Task 5                    95.93          3.00            96.60                           91.80          5.11        93.30

Task 6                    96.64          3.26            96.60                           93.80          4.87        93.30

__________________________________________________________________________________________

Table 7: Mean, SD and Median for accuracy for Group I (High proficient) and Group II (Low proficient) across the two tasks of cognitive flexibility

In addition to mean reaction time, accuracy scores was also computed for the two groups. Accuracy scores for the high proficient group (Mean = 95.93, Median = 96.60) was greater compared to the low proficient group (Mean = 91.80, Median = 93.30) on Task 5. Scores obtained on Task 6 also followed a similar pattern, (Mean = 96.64, Median =96.60) for high proficient group and (Mean = 93.80,  Median = 93.30 for low proficient group with the high proficient group obtaining better accuracy than the low proficient group. .

Mann Whitney U Test was carried out to verify if there was any significant difference between high and low proficient groups on the two tasks of cognitive flexibility. While measuring the reaction time on Task 5 the |Z| score obtained was 0.460 and the corresponding p value showed no significant difference. While measuring accuracy the |Z| score obtained was 2.999 and the corresponding p value showed significant difference. On Task 6 of cognitive flexibility the |Z| score obtained for mean reaction time was 0.027 and the p value indicated no significant difference , however for accuracy the |Z| value obtained was 1.941 and the corresponding p value showed a significant difference. In a nutshell there was a statistically significant difference observed on the accuracy scores for high and low proficiency groups on both Task 5 and Task 6 (p<0.05). However in terms of reaction time statistically significant difference was not seen across the groups (for both the tasks).  From all these tasks it can be inferred that the high proficient bilinguals outperformed the low proficient bilinguals

Bilingualism is defined as the usage and proficiency in at least two languages, which changes depending on the exposure to other users of the languages and the opportunities to use the languages. (ASHA, 2004). Since people tend to use more than one language in daily life. (Kroll & DeGroot, 2005) it can be seen that when a bilingual person uses one language, the other language is also active at the same time. For bilingual people, this activation is not limited to a single language as the auditory input activates corresponding words regardless of the language to which they belong.

The findings of the current study is in line with the research carried out in the past which have  found that bilingualism can enhance certain executive functions which include response inhibition, set shifting, attention and memory and the presence of a bilingual advantage being present. To maintain the relative balance between two languages, the bilingual brain relies on executive functions, because both of a bilingual individual’s language systems are always active and competing, and hence they use these control mechanisms and it is this constant practice that strengthens the control mechanisms and changes the associated brain regions. (Marian & Shook, 2002).

The reason that bilinguals are able to selectively attend to important attributes of a stimulus has been explained by the fact that bilinguals have different representations in each language for similar concepts and hence they need to be aware of what word to use in what context and the language which they are using to carry out the task by inhibiting irrelevant information and attending to the relevant stimuli. Bilinguals also can use two languages in the same modality better as compared to monolinguals which require them to make the decision about how to respond in a situation and what language to use based on the context (Bialystok and Martin, 2008).

The findings are consistent with the study done by Peal and Lambert (1962) who reported of bilingual advantage being seen consistently in various domains of executive functioning , perception and development of language. Also better and faster sensory and cognitive processing was seen in bilinguals which was attributed to the fact that there is manifestation of a bilingual cognitive advantage where an individual can inhibit irrelevant information and focus on the relevant information.

In the current study bilinguals outperformed monolinguals on tasks of attention which is consistent with the findings from the past where researchers have found out that the cognitive advantage seen in bilinguals over monolinguals results from executive attention (Bialystok, 1999; Yang, 2004).

Also on tasks of response inhibition , bilinguals obtained faster responses to both congruent and incongruent stimuli as compared to monolinguals and shorter reaction times which can be attributed to the fact that they have better interference suppression because of the ability to manage attention by inhibiting distracting spatial cues in rapidly changing contexts (Rodrigues & Zimmer , 2016)

Findings from the cognitive flexibility domain of executive function indicated better performance by bilinguals than monolinguals suggesting that life-long experience in switching between languages leads to increased efficiency in the ability to shift flexibly between mental sets (Prior & MacWhinney , 2010)

Since it’s rare to get monolinguals in the current scenario as almost every individual would have some exposure to another language apart from their mother tongue, the current study has considered high proficient and low proficient bilingual population to see how proficiency has an effect on executive functioning. As seen on the measures of reaction time and accuracy the high proficient group outperformed the low proficient group thus suggesting the presence of a bilingual advantage and also confirming the fact that there is a relation between bilingual proficiency and executive functioning.

Objective 2:

Comparison between Task 1 and Task 2 for the three domains between the high proficient and low proficient group.

Two tasks were used to tap the each domain of executive functions, the two tasks varied in terms of the linguistic load associated with each of the task. The even tasks (task 2, 4 and 6) were more linguistically loaded as compared to the odd tasks (task 1, 3 and 5).

Figures mentioned below depict the performance of the high proficient and low proficient group in terms of mean reaction time and accuracy on the two tasks (odd and even)  across the three domains.

1. Attention

Figure 1: Mean scores for reaction time for group I and group II on Task 1 and Task 2

Figure 1 depicts the time taken for completing the two tasks.  Mean reaction time for Task 1 was found to be lesser (Mean = 651.56) compared to Task 2 (Mean = 712.96) in the high proficient group; whereas in the low proficient group the time taken to complete Task 1 (Mean = 745.86) was more compared to Task 2 (Mean = 704.19.

Figure 2: Mean scores for accuracy for group I and group II on Task 1 and Task 2

As shown in Figure 2 the accuracy of responses for Task 1 was greater (Mean = 96.44) than Task 2 (Mean = 92.96) in the high proficient group. Similar trend was also seen in the low proficient group as there was better accuracy of responses for Task 1 (Mean = 93.96) than Task 2 (Mean = 86.63)

Wilcoxn Signed Rank Test was carried out to verify if there was any significant difference between the two tasks of attention. On measuring the difference between the two tasks of attention in terms of reaction time the |Z| score obtained was 2.949 and while measuring accuracy, the |Z| score obtained was 2.859. The corresponding p values showed significant difference in both the domains in the high proficient group. Similar trend was seen for the low proficient group in the domain of reaction time the |Z| score obtained was 2.539 and while measuring accuracy the |Z| score obtained was 3.933 and the corresponding p values showed significant difference in both the domains.

In a nutshell, there was a uniform trend seen as statistically significant differences (p<0.05) were obtained on both the measures of reaction time and accuracy across the two tasks of attention. As the mean reaction time was more and accuracy was less for even task, it can be inferred that the even task was more complex compared to the odd task for the attention domain.

2. Response inhibition

Figure 3: Mean scores for reaction time for group I and group II on Task 3 and Task 4

As depicted in Figure 3 the time taken for completing Task 3 was lesser (Mean = 743.22) than Task 4 (Mean = 1089.97) in the high proficient group. Similarly the low proficient group also took lesser time to complete Task 3 (Mean = 719.60) than Task 4 (Mean = 1120.75).

Figure 4: Mean scores for accuracy for group I and group II on Task 3 and Task 4

With respect to Figure 4 it can be seen that the accuracy of responses for Task 3 was almost equivalent (Mean = 92.29) to that of Task 4 (Mean = 92.31) in the high proficient group. However in the low proficient group there was better accuracy of responses for Task 3 (Mean = 91.46) than Task 4 (Mean = 87.30)

In order to verify if there was any significant difference between the two tasks of response inhibition, Wilcoxn Signed Rank Test was carried out. On measuring the difference between the two tasks terms of reaction time the |Z| score obtained was 3.920 and the corresponding p value showed significant difference. However while measuring accuracy the |Z| score obtained was 0.862 and the corresponding p values showed no significant difference in the high proficient group. In the low proficient group in the domain of reaction time the |Z| score obtained was 3.920 and while measuring accuracy the |Z| score obtained was 2.709 and the corresponding p values showed significant difference in both the domains.

To summate it was observed that statistically significant differences (p<0.05) were obtained on the reaction time scores on both Task 3 and Task 4 of response inhibition. Similar trend was seen for accuracy scores on Task 4. However the accuracy scores of Task 3 indicated no significant differences (p>0.05).  Based on the numerical findings it can be inferred that the odd task was more simpler compared to even task especially for low proficient group. 

1. Cognitive flexibility

Figure 5: Mean scores for reaction time for group I and group II on Task 5 and Task 6

As depicted in Figure 5 it can be observed that the time taken for completing Task 5 was in par (Mean = 730.81) with the time taken for completing Task 6 (Mean = 730.39) in the high proficient group whereas in the low proficient group the time taken to complete Task 5 (Mean = 719.42) was lesser than Task 6 (Mean = 730.17)

Figure 6: Mean scores for accuracy for group I and group II on Task 5 and Task 6

Figure 6 shows the accuracy of responses for the two tasks of cognitive flexibility. In the high proficient group Task 5 (Mean = 95.93) had slightly lesser accuracy of responses when compared to Task 6 (Mean = 96.64). Similar trend was also seen in the low proficient group as there was better accuracy of responses for Task 6 (Mean = 93.80) than Task 5 (Mean = 91.80)

Statistical analysis was carried out to verify if there was any significant difference between the two tasks of cognitive flexibility. Wilcoxn Signed Rank Test was administered.  On measuring the difference between the two tasks in regard to reaction time, the |Z| score obtained was 0.635 and on measuring accuracy the |Z| score obtained was 1.025 and the corresponding p values showed no significant difference in the high proficient group. In the low proficient group in the domain of reaction time the |Z| score obtained was 0.784 and the |Z| score obtained for accuracy was 1.206 and the corresponding p values showed no significant difference for both the domains.

In summary while comparing across mean reaction time and accuracy no statistically significant differences (p>0.05) were observed across both the tasks of cognitive flexibility. It can be inferred that the performance did not vary as a function of task complexity for the cognitive flexibility domain. 

The overall findings in the domains of attention and response inhibition are suggestive of the fact that the participants took more time to complete the even tasks which is consistent with the results depicted above. The even tasks being more linguistically complex took more time to be completed when compared to the odd tasks. This finding is in consonance with the study conducted by Bialystok et al. (2004) who reported of better performance on tasks which were relatively simpler. However as the complexity of the task was increased it was reported that the participants took longer time to perform the tasks but the presence of a bilingual advantage still persisted.   Another study in support of this finding was Kramer’s (2011) study where he found out that there may be an absence of literacy practices or low proficiency in several bilinguals and due to such inconsistencies it maybe relatively easier for a bilingual to perform less complex tasks better. Also Bialystok (2009) concluded that bilingualism is a very important factor in cognitive performance of an individual. However for the cognitive flexibility domain the performance did not vary as a function of task complexity which can be supported by the claim that under high processing demands the efficiency of the executive network was  more evident and hence the performance across the two tasks were similar. (Costa et al, 2009)

iv. Comparisons across the alternating fluency tasks between the high proficient and low proficient group.

Under each lexical category, the person had to name a lexical item in Kannada and name one lexical item in English. The participant was instructed to adhere to the lexical category and name different lexical items in both the language (not translation equivalents). The number of correct lexical items named in L1 (Kannada) and L2 (English) mentioned under each lexical category was taken into consideration and quantitative analysis was carried out. The overall mean, standard deviation (SD) and median were calculated for the performance of Group I (High proficient) and Group II (Low proficient) across the three categories.

Table 4.7 shows the performance of both the groups across the three categories.

__________________________________________________________________________________________

                                 Mean         SD          Median                          Mean        SD       Median

L 1                           14.25          3.79          15.00                           13.75        3.64        14.00

L 2                           19.40          5.29           20.00                           16.45        5.47       17.00

L 3                           12.85          4.22           13.00                            11.80       3.33       12.00

__________________________________________________________________________________________

Table 8: Mean, SD and Median for Group I (High proficient) and Group II (Low proficient) across the three categories of alternating fluency.

Note: L1, L2 and L3 refers to the three lexical categories of fruits, animals and vehicles.

As depicted in Table 8 the mean score obtained by the high proficient group (Mean= 14.25) was greater than the low proficient group (Mean= 13.75) on Category 1. Similar trend was also seen on Category 2 (Mean = 19.40 for high proficient group and Mean = 16.45 for low proficient group) and also for Category 3 (Mean = 12.85 for high proficient and Mean = 11.80 for low proficient) on the domain of alternating fluency. The median values of all the tasks followed the same direction.

Mann Whitney U Test was carried out to verify if there was any significant difference between high and low proficient groups across the three categories of alternating fluency. On category 1 the |Z| score obtained was 0.694 and the corresponding p value showed no significant difference. Similar trend was seen in category 2 where the |Z| score obtained was 1.642 and in category 3 where the |Z| score obtained was 0.679 and the p values in both the categories indicated no significant differences. This task showed marked difference between high and low proficient groups.

Friedman test was carried out to see if there was a significant difference between the three categories. The results revealed χ² (2) = 46.50 , p< 0.01 Chi square value obtained was 46.50 for 2 degrees of freedom and the p value indicated significant difference (p<0.01). Hence Wilcoxn signed rank test was performed to verify if there was a significant difference between the three categories of alternating fluency. In the high proficient group on comparing between category 1 and category 2, the |Z| score obtained was 3.551 and the corresponding p value showed significant difference. Similar trend was seen while comparing category 2 and category 3, the |Z| value obtained was 3.834 and p value indicated significant difference. However while comparing category 1 and category 3 the |Z| score obtained was 1.525 and the corresponding p values showed no significant difference.

In the low proficient group on comparing between category 1 and category 2, the |Z| score obtained was 2.278 and the corresponding p value showed significant difference. Similar trend was seen while comparing category 2 and category 3, the |Z| value obtained was 3.634 and while comparing category 1 and category 3 the |Z| score obtained was 2.530 and the corresponding p values showed significant differences.

To summate comparisons on the alternating fluency task between category 1 and 2 and between category 2 and 3 revealed statistically significant differences in the high proficient group. The same trend was seen in the low proficient group when comparing between each of the three categories. However statistically significant difference was not observed on the comparison between category 1 and 3 in the high proficient group.

The findings suggested that there was a better performance by the high proficient group across the three categories of animals, fruits and vehicles. Since the alternating fluency task is a cognitively loaded task that utilizes the cognitive check mechanism, the findings were also conclusive of the fact that the high proficient group performed better than the low proficient group as they were cognitively more efficient in switching across the two languages. Another support is derived from Hommel et al (2011) who claimed that mastering more than one language helps an individual to achieve a focused cognitive-control state that exerts a strong impact for processing information.

Overall the results of the present study revealed differences between the high proficient and the low proficient groups in the measures of reaction time and accuracy on the various domains of executive functioning namely attention, response inhibition and cognitive flexibility with the high proficient group having a superior performance. Also the study explored the differences between the two groups across the two tasks (odd and even tasks) which were used to tap each of the three domains which revealed that the odd tasks were relatively simpler and less linguistically loaded when compared to the even tasks. Finally the groups were compared on the alternating fluency task which also revealed better performance by the high proficient group.

Conclusions:

To conclude the study helps us to understand the relationship between executive functioning and bilingualism with the high proficient group outperforming the low proficient group across all the three domains. The study also found a relationship between the language load and executive functions. The alternating fluency task showed a superior performance by the high proficient group as they were cognitively more efficient in switching across the two languages. The executive functions varied as a function of degree of bilingual proficiency. This again supports the claim that bilingualism is advantageous and the advantages of bilingualism would scatter through a variety of executive functions.