Kannada Hd Picture

[Zee Petty]

Thisstudy presents normative data in Kannada for 180 coloured Snodgrass & Vanderwart pictures. Data are presented for naming latency, image agreement, picture-name agreement, familiarity, visual complexity, and age of acquisition (AoA). Sixty-eight native Kannada speaking adults completed all tasks. The effects of the rated variables on naming latency were examined and compared with data on the same variables in other languages. A regression analysis revealed that image agreement, name agreement, familiarity, and age of acquisition all had a significant impact on naming latency, while visual complexity and frequency did not. Although, the correlations among rated variables in Kannada were equivalent to previous normative studies, the cross-linguistic comparison revealed that only AoA was strongly correlated with other studies. The findings point to the importance of understanding the interplay of psycholinguistic variables on naming latency in different languages.

Copyright: 2022 Bangalore et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The normative data can be found here as supplementary material Appendix. It is attached as hyperlink at the end of manuscript in supporting information after references.

Picture naming is widely used in speech production research to address psycholinguistic, neuropsychological, and bi/multilingualism questions [1]. One of the most widely used sets of object-naming stimuli were produced by Snodgrass and Vanderwart [2]. Their set comprises 260-line drawings with norms for four psycholinguistic variables shown to operate during picture naming provided by a total of 219 American English speakers. Snodgrass and Vanderwart [2] split their participants into groups to collect ratings on the four psycholinguistic variables: image agreement (N = 42), name agreement (N = 40), familiarity (N = 40), and visual complexity (N = 40). The other 57 participants completed two additional image agreement tasks: Picture-Name agreement task (N = 40) and Image Variability rating (N = 17). These variables have been extensively investigated in multiple subsequent studies, with a colourized version of the images produced in 2004 [3].

Whilst these findings have been widely reported, there is growing recognition of limitations of generalizing norms across languages [24]. First, cultural differences may result in some items being unknown in some parts of the world. For example, artichoke/asparagus/celery/chisel/nut/pliers/plug/raccoon/wagon/wrench/French horn/ were not known in Russian [15]. Other items may have very different frequency of occurrence in different languages such as harp/accordion/trumpet/flute which are included in English norms but have low familiarity in Turkish [20] for example. Second, images of items with low familiarity in any given culture may be perceived differently: e.g., /cigarette/ may be perceived as /pencil/. Third, some items may not have a specific name in a language and are known by their English names across cultures such as /penguin/, /stapler/ etc. Fourth, relationships between conceptual representations and lexical ones differ across languages, and an image which evokes a specific and single name in English might generate a more general name or multiple names in another language. For example, /bird/ is referred to as /pakshi/ or /hakki/ in Kannada, a Dravidian language spoken by at least 44 million [25] people in Karnataka in the Southwest of India, as well as linguistic minorities in Maharashtra, Andhra Pradesh, Tamil Nadu, Telangana, Kerala.

Each participant completed the Modified Language Proficiency Questionnaire (MLPQ; see S1 Appendix). Twelve of the 68 participants were bilinguals, and the rest were multilinguals. All were native Kannada (L1) speakers of whom 66 had English (L2) as their second language and Hindi as their third language. The other two participants spoke a different second language and had English as their third language. All participants were late bi-/multilinguals, which means they were exposed to one language at birth (Kannada) and to another language or languages (English and/or Hindi) later in childhood or adulthood [38]. English was acquired later among the OA group (12+ years) than the YA group (7+ years).

The study received ethical approval from the School of Psychology and Clinical Language Sciences Research Ethics Committee at the University of Reading (2019-072-AA). The information sheet was provided in Kannada to each participant, and they were given the opportunity to ask questions about what the study would involve. They were also informed that they were free to withdraw from the study at any point in time without giving any reason. Following this, they were asked to provide their written consent.

In the present study, both the tasks and order of completion, were designed to collect data from the same participants for: IA, PNA, FAM, VC, AoA and naming. Name agreement scores were calculated from confrontation naming. Although normative and confrontation naming was conducted on the same day, the stimuli used for both were different.

In the confrontation naming task, items were presented one at a time using Psychopy software version 3.1.2 [39]. Participants were asked to name each picture as soon as it appeared on the computer screen. The task comprised 3 blocks. The first block consisted of five practice trials to ensure the participant understood the task efficiently. The second and third blocks were the main task and included 45 items each, with a five-minute break provided between each block. Each trial consisted of a 500ms fixation followed by the presentation of the picture item. A 200ms beep was presented simultaneously with the picture stimuli, which acted as the cue for measuring reaction time. The picture remained on the screen for 3000ms followed by a 2000ms blank screen.

In addition to the ratings collected from the participants, word frequency data for 162 items were obtained from Sketch engine Kannada web 2012 (KNWAC12) database (approximate word count: 11 million). Frequency data were log transformed (Log10 (Counts per million) + 1) and included as a predictor of naming latency in a regression analysis alongside the rated variables.

The overall mean and standard deviation on MoCA and MLPQ scores of the 68 participants are presented in Table 2. The MoCA scores are slightly but not significantly higher for the younger adults and the MLPQ scores indicate high L1 proficiency in both groups with above average proficiency for L2.

Item by item rating and name agreement data can be found in the S3 Appendix. The first step in examining the data was to establish whether to keep the data from the younger and older adults separate or to combine into one set. To address this, correlations were run between the YA and OA mean rating scores for each item for all psycholinguistic variables plus the confrontation naming results (see Table 3). There were strong significant positive correlations for all rated psycholinguistic variables and a moderate correlation for naming reaction time, although all were significant at p The results of the rating tasks, name agreement and reaction time for the 180 items are contained in Table 4. The mean scores are high for IA and PNA, indicating that the Kannada translation was representative of the image and vice-versa. Similarly, higher FAM and VC mean scores representative of the concepts were familiar and there was sufficient detail to identify the concepts. Low mean rating scores were obtained on average for AoA indicating the concepts were acquired early in life (the majority before 6 years). The results of the rating tasks are explored separately before examining their combined influence on naming latency.

Name agreement in the form of the Hstats was calculated from confrontation naming data for each item (Table 4). Hstats differs from Hpercent by considering all the naming responses given for an item even if it is not named in Kannada. On the other hand, Hpercent is the percentage of name agreement obtained for each item by calculating only the total number of correct responses (items correctly named in Kannada) provided by the total number of participants multiplied into 100 (see Table 4 for summary statistics). Reaction time measures for four items (Barn, Football, Lobster, and Sweater) were removed from the confrontation naming latency as these items had zero correct responses, accounting for 2.2% of the 180 items RT.

The Hstats scores for the 180 items demonstrated a range of name agreement from complete to very low name agreement across items. Forty-two items (24%) had Hstats of zero, meaning they had complete name agreement. Twenty of the 180 items had Hstats score above 1.5 indicating very low name agreement (arrow, barrel, bed, bird, boot, broom, cat, clothes pin, dress, iron, knife, owl, pear, socks, stove, umbrella, vase, wagon, watering can, zebra) with multiple alternative naming responses. The remaining 118/180 items represent Hstats score below 1.5 suggesting more than one name was being used by participants to represent these items.

To calculate the mean reaction time for each item name, incorrect responses and naming failures were first removed, which accounted for 36% of the responses. Naming latencies more than three standard deviations from the mean were excluded to reduce the effect of outliers which resulted in 1.13% of total responses. Naming latencies for all 176 items are reported in S3 Appendix. The overall mean naming latency for 176 coloured S&V pictures was 1634.78ms (SD = 313.78). Further, only correct responses were considered for the regression analysis and cross-linguistic correlations.

3a8082e126