Kannada akshara knowledge in primary school children: measurement of accuracy and reaction time using a cross-sectional study design

Introduction

Reading is a process of developing a sense of written material through systematic mapping of phonemes to the corresponding grapheme. Language-specific phonological and orthographic knowledge is crucial for successful word decoding. Orthographic knowledge is the information that is stored in memory, which dictates the rules to represent the spoken language in the written form (Apel, 2011). The role of orthographic knowledge is crucial in literacy acquisition (e.g., Apel et al., 2006; Castles & Coltheart, 2004). The studies conducted in alphabetical languages have reported that orthographic knowledge acts as the strongest predictor of reading success in children (Bowey, 2005; Caravolas et al., 2013; Hulme et al., 2012; Seymour, 2005). However, there are prevailing arguments about the generalization of theoretical models explaining the reading process and other research findings from alphabetic writing systems to children learning to read in other writing systems (Karanth, 2003; Nag, 2007). Studies in alphasyllabaries have shown that the developmental pathway of reading is different from children acquiring an alphabetic writing system (Nag, 2007; Nag & Snowling, 2011; Nag & Snowling, 2012). Alphasyllabaries are phonologically and orthographically different from alphabetic writing systems and share features of both alphabetic and syllabic writing systems. Previous studies have illustrated clear differences between alphabetic and alphasyllabaries languages (e.g., Nag, 2007; Nag, 2014; Tiwari et al., 2011). Alphasyllabaries represent sounds at the level of the syllable called akshara, but have distinct features to indicate sub-syllabic information (Bright, 1999). These differences explain the variability in the acquisition of orthographic knowledge across orthographies. For example, Seymour et al. (2003) reported that children acquired orthographic knowledge by the end of Grade I in alphabetical language such as English, whereas Shu (2003) reported it to be beyond Grade VI in Chinese.

Previous attempts have examined akshara knowledge in alphasyllabary languages, such as Kannada and Malayalam. Nag (2007) investigated children’s acquisition of orthographic knowledge in Kannada, a south Indian alphasyllabary language. The study included a group of 374 primary school children between the age range of 5 to 10 years, studying in regular schools with Kannada language as the medium of instruction. The children were assessed for akshara knowledge twice. The first assessment was done when they were Grade I, II, and III. The follow-up assessment was done after 15 months when the children moved to subsequent Grades. A Syllabograph Recognition Test (LAB, 2003) was conducted, which consists of 20 Kannada akshara, in which one akshara was a primary vowel, eight aksharas of consonants with inherent vowels, five aksharas of consonants with vowel diacritics, and six aksharas of consonant clusters. The results revealed a significant main effect of time of testing and Grade level with a significant interaction between these variables on akshara knowledge. Akshara mastery was found to continue into elementary school with an accuracy of around 80% during the second assessment. The study noted the differential trend for individual akshara types; consonants with inherent vowels were mastered by the end of Grade I, consonants with vowels diacritics with ligatures and consonant clusters were only partially mastered by Grade II, and this extended beyond Grade IV with a substantial number of children showing less than 50% accuracy even at Grade IV. These findings indicated that orthographic knowledge in Kannada was determined by the orthographic features of akshara.

Another study was conducted by Tiwari et al. (2011) in Malayalam, which is another alphasyllabary language. This study investigated akshara knowledge in Grade III children learning to read alphasyllabary. The study included 80 children between the age range of 7–8 years from two Malayalam medium schools in Kerala. The stimuli were similar to the Nag (2007) study, which included 17 Malayalam aksharas, such as a primary form of vowels, consonants with inherent vowels, consonants with vowels diacritics, and consonant clusters. The aksharas was presented in the flashcards one at a time, and the children were asked to name them. The results showed a differential pace of mastery of various akshara types in Malayalam alphasyllabary. Grade III children mastered vowels in primary form and consonants with inherent vowels. Consonants with vowel diacritics showed partial dispersion, whereas the consonant cluster showed extreme dispersion. They also reported that consonant with vowel diacritics was mastered much better than consonant clusters.

Nag (2007) and Tiwari et al. (2011) have indicated that the acquisition of akshara knowledge continues beyond Grade III in Malayalam and Grade IV in Kannada. However, there is a paucity of data beyond Grade IV in alphasyllabary languages, specifically in Kannada. Additionally, the studies in alphasyllabary have used a limited sample of aksharas out of a large registry, thus limiting the generalization of the findings. More importantly, the studies in alphasyllabaries have determined mastery by assessing the accuracy of akshara identification. However, fluent reading requires the speed with which the individual aksharas are retrieved from long-term memory. Hence, the speed with which akshara naming can be determined by measuring reaction time for individual aksharas. The speed along with accuracy in recognizing aksharas enables children to decode the words faster, thereby enhancing fluency in reading, resulting in improved comprehension. Hence the present study aimed to develop a comprehensive understanding of akshara knowledge acquisition by measuring both accuracy and reaction time in typically developing Kannada medium primary school children from Grade I to Grade VII within the age range of 5 years 6 months to 12 years 6 months.

Methods

Results

Figure 1 and Figure 2 show the mean and SD for akshara identification accuracy and reaction time across the Grades, respectively.

Figure 1. Mean and standard deviation for akshara identification accuracy from Grade I to Grade VII.

Figure 2. Mean and standard deviation for reaction time from Grade I to Grade VII.

The result of one-way ANOVA revealed a significant main effect of Grades on akshara identification accuracy [F (6, 308) = 1687.057, p=0.00] and reaction time [F (6, 308) = 69.84, p= 0.00]. Subsequently, a post-hoc pair-wise multiple group comparison was performed using the Bonferroni test, which revealed that the mean performance of Grade I on akshara identification accuracy was significantly lower than higher Grades (p≤0.05). This result was the same for Grades II, III, IV and V in comparison to higher Grades (p≤0.05). However, there was no significant difference between Grade VI and Grade VII (p≥0.05).

The post-hoc pair-wise multiple group comparison of mean reaction time using Bonferroni test failed to reveal any significant difference (p≥0.05) in mean performance between Grade I and II, but Grade I children took significantly more time (p≤0.05) in naming aksharas than all higher Grades. This result was the same for Grades II and III in comparison to all higher Grades (p≤0.05). However, there was no statistically significant difference between Grade IV and Grade V, but the arithmetical mean showed Grade IV took a longer time than Grade V. The performance of Grade IV was significantly lower than Grades VI and VII (p≤0.05). Similarly, there was no significant difference (p≥0.05) noted in the mean performance between Grades V and VII, but the mean performance of Grade V was significantly lower than Grade VII (p≤0.05). There was no significant difference noted (p≥0.05) in the mean performance between Grades VI and VII, which indicated that children from Grades I to IV took significantly greater time to identify the akshara compared to Grades VI and VII who took substantially less time.

Discussion

The results of one-way ANOVA revealed that accuracy in identifying akshara improves with Grade. On post-hoc pair-wise multiple group comparison, it was observed that the mean performance of children in identifying Kannada aksharas improved significantly from Grade I through to Grade V and reached a plateau at Grade VI, indicating a steady improvement in children’s akshara identification scores. The results showed a developmental trajectory, where children mastered every level of representative aksharas gradually in each Grade, and completely acquired all levels of representative aksharas by Grade VI. These findings are in agreement with results obtained by Nag (2007), who reported that the acquisition of akshara knowledge goes beyond ten years of age (Grade IV). In alphabetical languages, like English, children are reported to master letter knowledge by Grade I (Seymour et al., 2003) due to the lower number of graphemes compared to alphasyllabary languages, which consists of an extensive number of graphemes and also requires the mastery of a large group of orthographic register and their ligaturing rules (Nag, 2007). In alphabetic knowledge, the acquisition of letters of the alphabet takes place in a gradual manner, i.e. children learn the letters of the alphabet on the letter by letter basis, rather than learning the entire set at once. Therefore, young children are introduced initially to consonants followed by vowels, as consonants are more consistent in their sounds and hence can be quickly learned compared to vowels (Strickland, 1998). The gradual, steady improvement in the identification of akshara can be attributed to the manner in which aksharas are taught in schools. Generally, aksharas are taught to children in a hierarchical manner where they are introduced to the aksharas based on the level of complexity, i.e. initially, children are introduced to primary vowels and consonant with inherent vowels, followed by consonants with vowel diacritics, and consonant clusters (Karanth & Prakash, 1996).

The findings of reaction time measurement from the present study revealed that the children from Grade I to IV took a significantly greater time to identify aksharas compared to Grades V, VI and VII, who took significantly less time. This implies that younger children took a significantly longer time in identifying aksharas compared to older groups. The reduction in reaction time can be described as developmental changes in the speed of information processing (Kail & Salthouse, 1994; Kail, 2008). Studies report that increases in the speed of processing are correlated with age-related improvements in performance on a variety of cognitive tasks, including reading, memory, solving arithmetic problems, finding their way, and reasoning (Kail, 2004; Kail, 2008). Even though there was an increase with Grades, the performance of reaction time reached a plateau at Grade IV, and there was no significant difference noted in time taken to name aksharas above Grade IV. The occurrence of plateau in reaction time can be attributed to familiarity of the curricular material to which children are exposed frequently during primary school years. For example, most of the text introduced to Grade I children included words which are formed by primary vowels and consonant with inherent vowels, while in older grades, children are introduced with the words with diacritic markers and clusters, thus making them more familiar to those levels of complexity. As a result, children are more familiar with the aksharas which makes naming speech reach a constant level, but exploring and learning new aksharas continues across grades. Thus, automaticity in naming aksharas develops prior to akshara identification accuracy, which continues to develop above Grade IV and is mastered by Grade VI.

Conclusion

Children learning to read alphasyllabary Kannada language gain mastery over majority of the aksharas during the first five years of formal schooling and this develops completely by Grade VI. The automaticity in naming akshara develops gradually and reaches a plateau by Grade IV itself. The present findings indicate that children acquire automaticity in naming akshara early, whereas the akshara knowledge continues to develop.

Data availability