Development of knowledge test to measure the knowledge level of farmers on scientific pig farming practices

Introduction

Livestock play a vital role in sustaining the livelihoods of farmers in our country. Among various livestock species, pig farming serves as a major source of income for tribal communities, socially backward groups and economically weaker sections of society. It holds great potential due to several advantages, including high fecundity, prolificacy, a short generation interval, rapid body weight gain, efficient feed conversion and a superior dressing percentage compared to other livestock species (Das and Bujarbaruah, 2005 [3] ; Sastry and Thomas, 2021 [15] ). Despite these advantages, pig farmers in India, including those in Kerala, face several challenges, such as the unavailability of improved breeds, lack of knowledge on balanced ration feeding, frequent disease outbreaks and inadequate marketing facilities (Islam et al., 2016 [8] ; Pooja et al., 2024 [12] ). The success of piggery development programmes and the adoption of scientific pig farming technologies largely depend on the knowledge level of pig farmers. Therefore, assessing the existing knowledge levels of pig farmers is crucial for the effective implementation of welfare schemes and the dissemination of modern technologies. According to Bloom et al. (1956) [2] knowledge refers to includes those behaviours and test situations which emphasize the remembering either by recognition or recall of ideas, materials or phenomenon. Knowledge is a key component of behaviour and plays an important role in the covert and overt behaviour of an individual. A test is a set of questions, each of which has a correct answer, to which people respond (Ray and Mondal, 2011) [13] . A wellconstructed knowledge test helps determine the level of relevant knowledge of the target population. It indicates the extent of knowledge, with respect to scientific pig farming practices, that the respondents possessed at the time of the interview, as evident from their responses to a set of www.extensionjournal.com questions scientifically prepared for this purpose.

Materials and Methods

The present study aimed to develop and standardize a suitable knowledge test to assess the knowledge level of farmers regarding scientific pig farming practices, by following the procedure suggested by Davis (1951) [4] , Sagar (1983) [14] and Arun (2022) [1] which included the following steps.

Item collection

The content of a knowledge test consists of questions known as items. These items were prepared by referring to the Kerala Veterinary and Animal Sciences University Package of Practices Recommendations 2022, relevant textbooks and through discussions with experts from the Centre for Pig Production and Research, Mannuthy. Finally, a thorough scrutiny of the item pool was done with the assistants of subject matter specialist. The questions were designed to test the knowledge level of pig farmers regarding scientific pig farming. The items covered various domains such as scientific breeding, feeding, housing, health care, general management, piglet management and marketing practices.

Initial selection of items

The selection of items was done based on the following criteria.

• Items should promote thinking rather than rote memorization. • Items should differentiate the well-informed respondents from the poorly informed ones. • Items should have a certain difficulty value.

A total of 97 knowledge items were initially prepared.

Relevancy rating by experts

These prepared 97 items were subjected to scrutiny by an expert panel of judges to determine the relevancy and screening for inclusion in the final scale. The judges comprised faculties Extension Education, Livestock Production Management and pig farming experts from various departments across veterinary universities in India, as well as Assistant Directors working in the Department of Animal Husbandry, Government of Kerala. The items were sent to 98 Judges with request to critically evaluate each item for its relevancy to measure knowledge level of pig farmers in scientific pig farming practices. The judges were requested to give their response on a four-point continuum viz., very relevant, relevant, somewhat relevant and irrelevant with scores of four, three, two and one respectively. A total of 60 judges returned their responses, which were used for the relevancy analysis. The relevancy score of each item was established by adding the scores on the rating scale based on each judge's response. From the data, three types of tests relevancy percentage, relevancy weightage and mean relevancy scores were worked out for all the items. The items which had relevancy percentage 85 and above, mean relevancy weightage 0.85 and above, mean relevancy score 3.40 and above were considered for selection. Based on the above criteria a total of 55 items were selected for item analysis.

Item analysis

Item analysis is a technique through which those items which are valid and suited to the purpose are selected and the rest are either eliminated or modified to suit the purpose (Singh, 2006) [16] . It examines respondent's answers to individual test items (questions) to assess the quality of those items and the test as a whole. The procedure for item analysis was based on a very simple method for what happens when any person encounters any item. The method says that the outcome of such an encounter is governed by the product of the ability of the person and the easiness of the item. The more able the person, the better his chances for success with any item. The easier the item, the more likely that any person to solve it (Wright and Panchapakesan, 1969) [20] . A schedule was prepared with 55 items in different formats such as multiple-choice questions, true or false questions and open-ended questions. These items were administered to non-sampling 30 pig farmers across five districts of Kerala, viz., Palakkad (10), Idukki (6), Ernakulam (6), Thrissur (4) and Wayanad (4). All the items collected for the construction of the knowledge test were in the objective form. For each correct answer, one mark was assigned and while wrong answer was scored as zero. The total knowledge score of each respondent was calculated by adding the scores of all the questions. These knowledge scores were then used to calculate difficulty index, discrimination index and point bi-serial correlation which formed the criteria for selection of the items for the final knowledge test.

Difficulty index

The difficulty index of an item was defined as the proportion of respondents giving correct answer to that particular item (Ray and Mondal, 2011) [13] . An item should not be so easy that all respondents answer it correctly nor so difficult that no one can answer it. The difficulty index ranges from 0 to 100 per cent, with higher values indicating easier items. The recommended range of difficulty is between 25 and 75 per cent. Items with difficulty indices below 25 per cent were considered too difficult, while those above 75 per cent were considered too easy (Hingorjo and Jaleel, 2012) [7] . The difficulty index of each of the 55 items were calculated by dividing the total correct answers for a particular item by the total number of respondents. It was calculated by the following formula:

Where, Pi= Difficulty index in percentage of the i th item. ni =Number of respondents giving a correct answer to the i th item. Ni = Total number of respondents to whom the items were administered (30 in the present study) Items with values ranging between 25 and 75 per cent were considered for selection in the final knowledge test.

Discrimination index

The item discrimination index is the ability of an item to www.extensionjournal.com differentiate between respondents of higher and lower abilities (Namdeo and Sahoo, 2016) [11] . The statement which was answered correctly by everyone or the one which was not answered by anyone in the sample had no discrimination value. Therefore, only those statements with high power to discriminate the respondents who varied in the level of knowledge were included in the final list. The discrimination power of all the 55 items was worked out using E 1/3 method to find out the item discrimination, as given below. In this study, after calculating the total score obtained by 30 respondents, then the scores were arranged in a descending order. These respondents were divided into six equal groups, each having five respondents. These groups were named as G1, G2, G3, G4, G5 and G6 respectively. For item analysis, the middle two groups (G3 and G4) were eliminated. Only four extreme groups with high scores (G1 and G2) and low scores (G5 and G6) were considered to calculate the 'Discrimination index'. It is calculated by the following formula:

Where, N = Total number of respondents to whom the items were administered (30 in the present study). S1 and S2 = Frequencies of correct answers of highest and higher scores, respectively S5 and S6 = Frequencies of correct answers of lower and lowest scores, respectively.

Point bi-serial correlation

A correlation between a continuous and a dichotomous variable is known as the point-bi-serial correlation (Demirtas and Hedeker, 2016) [5] . It is a special case of the Pearson product moment correlation. It is used to calculate the correlation between the scores on a single test item (right or wrong) with scores on the total test (Vercruyssen and Hendrick, 2012) [19] . For establishing internal consistency of each item, the point biserial correlation coefficient (rpbis) was estimated. It is calculated by the following formula: Point bi-serial correlation was computed as where, rpbis = Point bi-serial correlation Mp = The mean of the total score of the respondents who answered an item correctly Mq = The mean of the total score of the respondents who answered an item incorretly σ = The standard deviation of the entire sample p = The proportion of the respondents giving correct answer to an item q = The proportion of the respondents giving incorrect answer to an item.

Results and Discussion

Final selection of the items for the test All the items which had a difficulty index ranging from 25 to 75 per cent, discrimination index ranging from 0.30 to 0.70 and point bi-serial correlation value that was significant at 5 per cent level were selected for the final knowledge test. Out of 55 items 18 items were finally selected.

Standardization of the scale Validity of the knowledge test

Validity is the ability of an instrument to measure what it is designed to measure (Kumar, 2014) [10] . The validity of the knowledge test was established through content validity.

The content validity of the knowledge test was ensured by choosing items in consultation with various subject matter specialists. All possible care was taken while selecting the items and the same was subjected to difficulty and discrimination index and point bi-serial correlation, to select the final items. Hence, it was logical to assume that the test satisfied representative approach of test construction, the criteria for content validity.

Reliability of the knowledge test

According to the Kerlinger (1973) [9] 'reliability' is the accuracy or precision of a measuring instrument and it is also concerned with repeatability or consistency. A measure is considered reliable if it gives the same result over and over again (Singh, 2013) [17] . An instrument cannot be valid unless it is reliable. The reliability of the knowledge test was assessed by calculating the Cronbach's alpha coefficient of reliability, which ranges between 0 and 1 (Singh, 2006 [16] ; Tavakol and Dennick, 2011 [18] ). For this purpose, the knowledge test was administered to 30 non-sample pig farmers selected randomly from Ernakulam (8), Idukki (8), Thrissur (5), Wayanad (5) and Thiruvananthapuram (4) districts of Kerala. The collected data were tabulated and analysed to estimate the alpha value. The formula for calculating the alpha value is as follows. www.extensionjournal.com

Where, α = Cronbach's alpha reliability coefficient. K = Number of items.

σ 2 yi = The variance of item i for the current sample of persons. σ 2 x = The variance of the observed total test scores.

Reliability statistics Cronbach's Alpha

No of items 0. 826 18 In the present study, the knowledge test yielded a Cronbach's alpha of 0.826 across 18 items, indicating a good level of internal consistency. This means that respondents who gave higher scores on one item generally gave higher scores on other items, and vice versa. The column containing the "corrected item-total correlation" indicates the correlation between a given item and the summated score of all the other remaining items.

For example, the correlation between item 1 and the summated score of items 2 to 18 was r = 0.620, indicating a positive association between the score on item 1 and the combined score of the remaining items (item 2 to item 18). The column containing the "Cronbach's alpha if item deleted" indicates the value of Cronbach's alpha that would result if an individual item was removed from the scale. For example, if item 1 was deleted, the Cronbach's alpha would decrease from 0.826 to 0.806, showing that item 1 contributes positively to the overall reliability of the knowledge test.

The rule of thumb for Cronbach's alpha (α) is as follows: α > 0.90 -excellent, α > 0.80 -good, α > 0.70 -acceptable, α > 0.60 -questionable, α > 0.50 -poor, and α < 0.50unacceptable (George and Mallery, 2003). In the present study, the knowledge test had a Cronbach's alpha value of 0.826, indicating a good level of internal consistency among the 18 items. Hence, it was concluded that the items used in the scale were both appropriate and reliable.

Administration of the knowledge test

The developed knowledge test will be administered to the respondents to assess their knowledge level regarding scientific pig farming practices. Each respondent will be given a score of 'one' for every correct answer and 'zero' for each incorrect response. The total knowledge score of each respondent will be calculated by summing the scores of all correctly answered items. The knowledge score for an individual will range from a minimum of 0 to a maximum of 18. Based on the total scores obtained, the respondents will be categorized into three groups viz., low, medium and high knowledge levels.

Conclusion

This study is a part of Ph.D. research work wherein a knowledge test was developed and standardized by using reliability and validity. The finalized test comprised a total of 18 questions. It is intended to assess the existing knowledge level of pig farmers regarding scientific pig farming practices. Identifying knowledge gaps through this tool will be highly useful for extension functionaries, policymakers and administrators in formulating skill oriented and need-based training programmes as well as implementing effective extension strategies for improving the livelihoods of farmers.

Acknowledgement

AcknowledgementsThis publication is part of the Ph.D. research work and we are greatly thankful to Kerala Veterinary and Animal Sciences University for their support and to the pig farmers of Kerala for their valuable participation.

References

1. Arun L. Livelihood security of dairy farmers affected by Kerala flood 2018. PhD thesis. Kerala: Kerala Veterinary and Animal Sciences University; 2022. p. 262.
2. Bloom BS, Engelhart MD, Furst EJ, Hill WH, Krathwohl DR. Taxonomy of educational objectives: handbook I, the cognitive domain. New York: Longmans Green and Company Ltd; 1956. p. 207.
3. Das A, Bujarbaruah KM. Pig for meat production. Indian J Anim Sci. 2005;75(12):1448-1452.
4. Davis FB. selection techniques. In: Lindquist EF, editor. Educational measurements. Washington (DC): American Council on Education; 1951. p. 266-328.
5. Demirtas H, Hedeker D. Computing the point-biserial correlation under any underlying continuous distribution. Commun Stat Simul Comput. 2016;45(8):2744-2751.
6. George D, Mallery P. SPSS for Windows step by step: a simple guide and reference. Version 11.0 update. 4th ed. Boston: Allyn & Bacon; 2003. p. 386.
7. Hingorjo MR, Jaleel F. Analysis of one-best MCQs: the difficulty index, discrimination index and distractor efficiency. J Pak Med Assoc. 2012;62(2):142-147.
8. Islam R, Nath P, Bharali A. Constraints perceived by the small-scale pig farmers in Sivasagar district of Assam: an analysis. Asian J Anim Sci. 2016;11(1):7377.
9. Kerlinger FN. Foundations of behavioral research. 2nd ed. Delhi: Surjeet Publications; 2004. p. 741.
10. Kumar R. Research methodology: a step-by-step guide for beginners. 4th ed. New Delhi: Sage Publications India Pvt Ltd; 2014. p. 399.
11. Namdeo SK, Sahoo B. Item analysis of multiple-choice questions from an assessment of medical students in Bhubaneswar, India. Int J Res Med Sci. 2016;4:17161719.
12. Pooja U, Balusami C, Murugan SS, Abraham J, George S, Pramod K. Evaluation of pig farmers' constraints in different agro-climatic zones of Kerala. Indian Vet J. 2024;101(12):15-19.
13. Ray GL, Mondal S. Research methods in social and extension education. 3 rd ed. Ludhiana: Kalyani Publishers; 2011. p. 330.
14. Sagar RL. Study of agro-economic, socio-psychological and extension communication variables related with farmers' productivity of major field crops in Haringhata block. PhD thesis. West Bengal: Bidhan Chandra Krishi Viswavidyalaya; 1983. p. 243.
15. Sastry NSR, Thomas CK. Livestock production management. 6th ed. Ludhiana: Kalyani Publishers; 2021. p. 1116.
16. Singh AK. Tests, measurements and research methods www.extensionjournal.com in behavioural sciences. 5th ed. New Delhi: Bharathi Bhawan Publications and Distributors; 2006. p. 602.
17. Singh RS. Research methods and statistical techniques. New Delhi: Anmol Publications Pvt Ltd; 2013. p. 352.
18. Tavakol M, Dennick R. Making sense of Cronbach's alpha. Int J Med Educ. 2011;2:53-55.
19. Vercruyssen M, Hendrick HW. Behavioral research and analysis: an introduction to statistics within the context of experimental design. 4th ed. Boca Raton: CRC Press, Taylor & Francis Group; 2012. p. 272.
20. Wright B, Panchapakesan NA. Procedure for samplefree item analysis. Educ Psychol Meas. 1969;29(1):2348.