The study was conducted at Bharabia and Dhamrai union of Dhamrai upazila under Dhaka district. Two villages Bharabia and Chandra Para of Bharabia Union and Asulia village of Dhamrai Union were purposively considered as the locale of the study because rice is intensively cultivated in these villages.

BR-29 rice variety cultivated by all farmers in the study area was the population of the study. According to Yamane’s [11] formula, the sample size was determined as 102. A reserve list of 10 farmers was prepared in case of their absence in any case. In calculating sample size 5% precision level, 50% degree of variability and value of Z = 2.57 at 99% confidence level were chosen from the following formula:

$n=\frac{z^{2}P\left(1-P\right)N}{z^{2}P\left(1-P\right)+N{\left(e\right)}^2}$

where:

n = Sample size, N = Population size, e = the level of precision and Z = the value of the standard normal variable at the chosen confidence level and P = the proportion or degree of variability.

Proportionate random sampling techniques were used to select a sample from three villages of the study area. According to the appropriate proportion of sample size data were collected from each village of the respective union. A reserve list of 10 farmers (about 10% of the sample) was kept purposively if any respondents were unavailable at the time of data collection. The distribution of population and sample size will be shown in Table 1 .

In order to collect valid and reliable information an interview schedule was prepared. It was carefully designed with the objectives of the study in mind. Both open and closed forms of questions were used to collect information. Simple, direct questions and scales were included in the interview schedule for collecting information regarding the focus on boro rice yield gap minimizing strategies. Interview schedules were pre-tested in actual field situations before being used for final data collection among 15 respondents in the study area. A reliability test was done. Necessary corrections, modifications and additions were made in the interview schedule on the basis of the results of the pre-test. The interview schedule was then printed in its final form. Necessary photocopies were then made.

Before data collection, the researcher met with the Upazila Agriculture Officer (UAO), Agriculture Extension Officer (AEO) and Sub-Assistant Agriculture Officer (SAAO) of those blocks for necessary help and cooperation. Data was collected in-person by the researcher himself through face to face interviews. The interview was conducted in the respondent’s farm and home during their leisure period. Before starting the interview, the researcher took all possible care to establish rapport so that the respondent did not hesitate to furnish proper responses to the questions and statements included in the interview schedule. However, if any respondent had difficulty understanding any questions, the researcher took the utmost care to explain and clarify the question.

In social research, the selection and measurement of variables constitute an important task. In this connection, the researcher looked into the literature to widen his understanding of the nature and scope of the variables involved in research studies. [12] defined a variable as any measurable characteristics which can assume varying of different successive individual cases. The hypothesis of a research, while constructed properly, contains at least two important elements, an independent variable and a dependent variable.

An independent variable is a factor that is manipulated by the researcher in an attempt to ascertain its relationship to an observed phenomenon [13] . A dependent variable is that factor which appears, disappears or varies as the experimenter introduces, removes or varies the independent variables. The dependent variables are often called the criterion or predicted variable, whereas the independent variable is called the treatment, experimental and antecedent variables [14] . Variables are very important for social research on which the statistical analysis will be done by the obtained scores on these variables. The eleven characteristics of the farmers were considered as independent variables of this study.

Age of a respondent is termed as the period of time from his birth to the time of interviewing. It was measured in terms of complete years.

The level of education of a respondent was measured on the basis of classes passed in a formal educational institution. For example, if a respondent passed up to class 4, his education score was taken as 4. If a respondent did not know how to read and write his education score was assigned as zero (0). A score of 0.5 will be given to that respondent who could sign his name only.

The farm size of a respondent was determined as the total area of his farm. It included as an area of the farm owned by him as well as those obtained from others by renting, leasing or other means.

Annual family income is referred as the annual gross income of a respondent family from agricultural production, business, service and income from other family members during the last one year. The income was expressed in Thousand Taka (‘000’ Tk).

The land area under boro rice cultivation of a respondent was measured by the total land area of his farm under boro cultivation last year. The land area under boro rice cultivation of a respondent was measured in hector.

Profit from boro rice cultivation refers to the annual net profit of a respondent from the production of boro rice. The profit from boro rice was expressed in Thousand Taka (‘000’ Tk). Profit from boro rice cultivation was estimated by subtracting the annual expenditure of boro rice cultivation from the income from boro rice cultivation.

The yield gap of boro rice of a respondent was referred to as the difference between the potential farm yield and the actual average farm yield. It was measured in ton/hector.

Training received by the respondent on rice cultivation was determined by the total number of days of training received by the respondent from any organization on rice cultivation. If a respondent took 2 days of training on any aspect of rice cultivation from any GOs, NGOs then his training received score would be 2.

Extension contact may be defined as one’s extent of exposure to different extension methods. The extent of contact was determined against four (4) point rating scales as not at all, rarely, occasionally and frequently and score was assigned as 0, 1, 2 and 3 respectively. The extension contact of a respondent was, therefore, determined by adding the total responses against 10 selected extension contacts. The extension contact score could range from 0 to 30, where 0 indicating no extension contact and 30 indicating very high contact.

Knowledge on rice yield gap minimizing strategies of a respondent was measured by using 12 different questions in relation to various rice yield gap minimizing strategies. It was measured in scores. A respondent was given full score ‘2’ for correct response. However, partial score was given for partially correct response and a ‘0’ score was given for wrong or no answer. The summation of score obtained by a respondent was the knowledge score of the respondent. The knowledge score of the respondents on rice yield gap minimizing strategies could range from 0 to 24 where ‘0’ indicating no knowledge on rice yield gap minimizing strategies and ‘24’ indicating the highest knowledge on rice yield gap minimizing strategies.

Ten (10) problems were selected and validated by experts to measure the extent of problem faced by the farmers in practicing rice yield gap minimizing strategies. Five (5) point rating scale was used for each problem. Five alternative responses were not at all, low, medium, high and very high problem. The weights were assigned to these responses as 0, 1, 2, 3 and 4 respectively. Extent of problem faced score of the respondents was measured by summing up all the responses to all the problems. The extent of problem faced score could range from 0 - 40 where ‘0’ indicating no problem and ‘40’ indicating very high problem.

Boro rice yield gap minimizing strategies practiced by the farmers were measured by rank order of the 12 selected boro rice yield gap minimizing strategies. The rank order of the selected boro rice yield gap minimizing strategies was done on the basis of mean value of practicing of each of the 12 strategies by the respondents.

Measurement of Extent of Practice of Boro Rice Yield Gap Minimizing Strategies

“Practice of boro rice yield gap minimizing strategies by the farmers” was the dependent variable of the study. For measuring the boro rice yield gap minimizing strategies practiced by the farmers, a five (5) point rating scale with 12 strategies were used. The 12 strategies were selected and validated by experts to measure the extent of boro rice yield gap minimizing strategies practiced by the farmers. Five alternative categories to measure the boro rice yield gap minimizing strategies practiced by the farmers were not at all, rarely, occasionally, often, regularly. The score assigned to these responses were 0, 1, 2, 3 and 4 respectively. The extent of boro rice yield gap minimizing strategies practiced by the farmers score ranged from 0 - 48, where ‘0’ indicating no practice and ‘48’ indicating highest level of practice of boro rice yield gap minimizing strategies. On the basis of extent of practice of boro rice yield gap minimizing strategies the respondents were classified into three categories.

The collected raw data were examined thoroughly to detect errors and omissions. As a matter of fact, the researcher made a careful scrutiny of the completed interview schedule to make sure that necessary data will be entered as complete as possible and well arranged to facilitate coding and tabulation. Very minor mistake was detected by doing this, which were corrected promptly.

After completion of field survey, all the data were coded, compiled and tabulated according to the objectives of the study. Local units were converted into standard units. All the individual response to questions of the interview schedule was transferred into a master sheet to facilitate tabulation and categorization.

The collected raw data as well as the respondents were classified into various categories to facilitate the description of the independent and dependent variables. These categories were developed for each of the variable by considering the nature of distribution of the data and extensive literature review. The procedure for categorization has been discussed while describing the variables under consideration in Chapter 3.

The data was analyzed in accordance with the objectives of the study. Qualitative data was converted into quantitative data by means of suitable scoring techniques wherever necessary. The statistical measures such as range, number, sum, mean, standard deviation, frequency, and percentage distribution were used for categorization and describing the variables. Multiple linear regression analysis was done to explore the contributing relationship between the selected characteristics of the farmers with the dependent variable. Statistical package for social sciences (SPSS) version 25 was used for the analysis of data. Five percent (0.05) level of probability was considered as the basis for rejecting any null hypothesis.

Eleven characteristics of the farmers were selected. The selected characteristics were age, educational qualification, farm size, annual family income, land area under boro rice cultivation, profit from boro rice, yield gap of boro rice, training received on rice cultivation, agricultural extension media contact, farmers’ knowledge on rice yield gap minimizing strategies and problem faced in practicing rice yield gap minimizing strategies. These characteristics of the farmers are described in this chapter. For ready reference, separate tables are provided while presenting categories, discussion and interpretations of results concerning each of the characteristics in this chapter.

Age

This distribution was supported by [15] and [16] shown in Table 2 , Data contained in Table 2 , indicate that an overwhelming majority (95.10%) of the farmers in the study area were middle to old aged. It may due to middle to old aged people have more land ownership than young aged people.

Education

This distribution was supported by [17] shown in Table 2 , Data presented in Table 2 , express that almost three fourth (74.50%) of the farmers had primary to secondary education, while 21.60% were illiterate and only 3.90% had above secondary level education. The authors of found almost similar findings. Most of the respondents of the locale had primary to secondary level of education and illiteracy rate was low. It may due to the area was adjacent to Dhaka, as well as primary schools and high school and college are located in the locale.

Farm size

Based on the farm size, the farmers were classified into three categories namely small farm size, medium farm size and high farm size as shown in Table 2 . Data contained in Table 2 , reveal that 70.59% of the farmers had small farm size, 21.26% of the farmers had medium farm size and 7.84% of the farmers had large farm size. The average farm size of the farmers of the study area (0.862 hectares) was higher than that of national average (0.60 hectare) of Bangladesh [4] .

Annual family income

Data presented in Table 2 , indicate that 74.50% of the farmers had low family income, 19.69% of the farmers had medium family income and 5.88% of the farmers had high family income. Dominance of low income farmers may due to poor socio-economic condition, small and medium farm size of the majority farmers. As well as mean annual income of the locale was lower than the national average of $1752 USD may due to more involvement of the farm families in business, services, and getting foreign remittance.

Land area under boro rice cultivation

Data presented in Table 2 , reveals that 41.17% of the farmers in the study area were small boro farmer, 42.16% were medium boro farmer and 14.70% were big boro farmer. Land area under boro rice cultivation shrinks form the total farm size of the respondents. It may due to distribution of farm size for various purpose e.g. homestead area, pond, cultivation of other crops, previous crops, potato, winter vegetables cultivation.

Profit form boro rice

Data contained in Table 2 , indicate that 35.29% of the farmers had low profit, 46.07% of the farmers had medium profit and 18.62% of the farmers had high profit. The majority of the farmers had medium to high profit which may be due to the high price of boro rice, availability of rice mills in the locale and good transport facilities with the capital.

Yield gap of boro rice

Data presented in Table 2 , revealed that an overwhelming majority (75.49%) of the farmers had a high yield gap and 18.62% of the farmers had a medium yield gap and 5.88% of the farm. High yield gap exists in the locale, which may be due to famer’s lack of knowledge, education, and low practices of rice yield gap minimizing strategies.

Training received on rice cultivation

Data contained in Table 2 , indicate that 43.10% of the farmers had low training exposure, 51.00% had medium training exposure and 5.90% had high training exposure on rice cultivation. Inadequate applied training facilities for the farmers to achieve high rice production may a reason behind low training exposure by majority of the boro rice farmers. Unwillingness of the farmers to receive and adopt training on modern rice cultivation practices may another reason behind this.

Agricultural extension media contact

Data presented in Table 2 , reveal that 23.53% of the farmers had low agricultural extension media contact, 68.63% of the farmers had medium agricultural extension media contact and 7.84% had high agricultural extension media contact. Extension contact of the respondents varied, which may be due to socio-economic conditions of the farmers. It was found that low income farmers had low extension media contact in the locale. Their involvement in day labour, small vendors and reluctance to agricultural extension media may the reasons behind this.

Knowledge on rice yield gap minimizing strategies

Data contained in Table 2 , reveal that 30.40% of the farmers had poor knowledge, 49.00% had moderate knowledge and 20.60% had good knowledge on rice yield gap minimizing strategies. Lack of education, agricultural extension media contact and training exposure to rice cultivation may the reason behind this.

Problem faced in practicing rice yield gap minimizing strategies

Data presented in Table 2 , reveal that an overwhelming majority (79.40%) of the farmers faced medium to high problem and 20.60% of the farmers faced low problem in practicing rice yield gap minimizing strategies. Farmers stated that high cost and unavailability of inputs, labor shortage, excessive cold during germination and transplanting period, pests and diseases infestation, excessive precipitation at harvesting period were the major problems faced by them.

Identification of boro rice yield gap minimizing strategies practiced by the farmers and ranking of boro rice yield gap minimizing strategies practiced are discussed in the following sub-sections. Boro rice yield gap minimizing strategies practiced by the farmers are ranked in ascending order on the basis of mean practice value as shown in Table 3 .

Data presented in Table 3 , revealed that strategies i.e., Sowing seedling within mid-December to mid-January; Transplanting 35 - 45 days old seedling; Maintaining sowing distance (20 – 25 × 15 − 20 cm) ranked 1^st with the mean of 4.0 and SD of 0. This happens because farmers in the study area are well informed about the impact of those strategies in rice yield gap minimization and they practice those strategies regularly.

Harvesting during maturity index (80% panicles have about 80% ripened spikelets) ranked 2^nd with the mean of 3.93 and SD of 0.254. This occurs because farmers of the study area are well aware of the maturity index and harvesting time of boro rice.

Use of IPM ranked 3^rd with a mean of 3.89 and SD of 0.312. This may be due to farmers of that area has sufficient knowledge on use of IPM and its impact on boro rice.

Following recommended steps of post-harvest operation (Using paddle thresher for threshing, sun drying 4 - 5 times, cleaning with winnower/air, store in gunny bags at 10 - 12% moisture content) ranked 4^th with the mean of 3.64 and SD of 0.541. This may be due to farmers of the study area had adequate knowledge on post-harvest operations of rice. But, due to lack of equipment for drying, threshing, storing and labour shortage yield gap may exist. Moreover, small farmers of the area do the operations manually by their own, it requires much time and cause grain deterioration.

Use of recommended irrigation (AWD, 7 - 10 cm transplanting to maximum tillering stage, from 40 DAT 12 - 15 cm up to late tiller production) ranked 5^th with the mean of 3.49 and SD of 0.502. The locale is under cover of irrigation facility and most of the farmers use AWD method of irrigation. It was found that, farmers having low income and lack of knowledge, awareness do not practice irrigation regularly which may be a reason behind rice yield gap in the study area.

Use of recommended weeding (15 DAT, 30/35 DAT and 45/50 DAT) ranked 6^th with the mean of 3.2 and SD of 0.423. Farmers of the study area has medium knowledge on weeding period of rice but due to labour shortage, large and medium farm size farmers sometimes cannot practice weeding operation in time. Moreover, in the present study it was found that most of the low-income farmers practice weeding manually by their own without labour, which requires much time, and it may be another reason behind rice yield gap in the locale.

Use of recommended doses of fertilizer (Urea-40 kg, TSP-13 kg, MoP-22 kg, Zipsum-15 kg, ZnSO₄-1.5 kg/ Bigha in 3 installment) ranked 7^th with the mean of 1.36 and SD of 0.842. This happened because farmers of the study area have lack of knowledge on fertilizer doses of boro rice to obtain potential yield and high cost of fertilizer. Moreover, residual impact and overdoses of fertilizer can reduce soil fertility and farmers had lack of knowledge regarding this issue.

Seed treatment with recommended fungicides (For funagal disease Agrosan, Bavistin, Vitavax 200, Homai 0.4% seed weight; Hot water treatment 54˚C for 15 min, Agrimycin 0.025% 12 hour before sowing for bacterial disease) ranked 8^th with the mean of 1.13 and SD of 1.04. In the present study, it was found that farmers of the study area had lack of knowledge on seed treatment and were unwilling to bear its expense. They often use traditional seed treatment method by boiling and soaking seeds 12 hour before sowing into the seedbed which is ineffective to hinder soil and seed born pathogens, which may ultimately cause outbreak of diseases.

Use of quality seed (TLS/ registered/ certified/ foundation seed) ranked 9^th with the mean of 1.12 and SD of 0.978. This may be due to most of the farmers regularly use their own seeds in rice cultivation. During data collection, it was found that, medium and big farmers of the locale occasionally use quality seed for rice production.

Use of crop calendar ranked 10^th the mean of 0 and SD of 0 because farmers of the study area had very few knowledge on rice crop calendar and its impact on minimizing rice yield gap Farmers of the locale do not use crop calendar.

Boro Rice yield gap minimizing strategies practiced by the farmers varied from 28 to 42 against the possible range of 0 - 48 with the mean of 33.75 and standard deviation of 2.996. Based on the extent of practice of boro rice yield gap minimizing strategies, farmers are classified into three categories (Mean ± SD) namely, “minimum strategies practiced”, “medium strategies practiced”, maximum strategies practiced” as shown in Table 4 .

Data contained in Table 4 , indicate that an overwhelming majority (90.19%) of the farmer practice low and medium level of boro rice yield gap minimizing strategies. Low annual income, low profit, small farm land, low training exposure, low agricultural extension media contact, lack of knowledge on rice yield gap minimizing strategies, high problem faced in rice cultivation and poor socio-economic conditions of the farmers may be the reasons behind minimum and medium level of strategies practiced by them.

This section deals with the findings exploring the contributing relationship between the selected characteristics of the farmers with their practice of boro rice yield gap minimizing strategies. The contributing factors were age, educational qualification, farm size, annual family income, land area under boro rice cultivation, profit from boro rice cultivation, yield gap of boro rice, training received on rice production, agricultural extension media contact, knowledge on rice yield gap minimizing strategies, problem faced in practicing rice yield gap minimizing strategies. The main focus of the study was, “Practice of boro rice yield gap minimizing strategies by the farmers”.

To assess the contributing relationship between selected characteristics of the farmers with their practice of boro rice yield gap minimizing strategies, a multiple linear regression analysis was done. The multiple linear regressions results have been shown in Table 5 .

The null hypothesis was, “Each of the eleven selected characteristics (age, educational qualification, farm size, annual family income, land area under boro rice cultivation, profit from boro rice, yield gap of boro rice, training received on rice cultivation, agricultural extension media contact, knowledge on boro rice yield gap minimizing strategies and problems faced in practicing boro rice yield gap minimizing strategies) of the farmers has no contributing factor with their practice of boro rice yield gap minimizing strategies.”

** Significant at p < 0.01, * Significant at p < 0.05.

The findings of the study revealed that, the eleven (11) characteristics of the farmers were taken as independent variables together were effective in predicting boro rice yield gap minimizing strategies practiced by the farmers. The observed F ratio was 68.861. Which was significant at 0.01 level of probability and indicated that the combination of the independent variables in boro rice yield gap minimizing strategies practiced by the farmers was effective? 83.8 percent (%) (R² = 0.838) of the variation in the respondents’ practice of boro rice yield gap minimizing strategies can be attributed to their age, educational qualification, farm size, annual family income, land area under boro rice cultivation, profit from boro rice cultivation, yield gap of boro rice, training received on rice production, agricultural extension media contact, knowledge on rice yield gap minimizing strategies, problem faced in practicing rice yield gap minimizing strategies making contribution on boro rice yield gap minimizing strategies practiced by the farmers.

However, each predictor may expound some of the variance in respondents’ practice of boro rice yield gap minimizing strategies. The adjusted R-square value penalizes the addition of external predictors in the model, but values of 0.825 still show that the variance in farmers’ practice of boro rice yield gap minimizing strategies attributed to the predictor variables rather than by chance and the F value indicate that the model was significant (p < 0.01).

From Table 5 , it was observed that farmers age, educational qualification, yield gap of boro rice, training received on rice production, agricultural extension media contact and problem faced in practicing rice yield gap minimizing strategies had significant contributing factor on boro rice yield gap minimizing strategies practiced by them. Data also showed that here educational qualification and yield gap of boro rice had most significant contributing factor at 1% (p < 0.01) level of significance on boro rice yield gap minimizing strategies practiced by the farmers. Moreover, the data showed that farmer’s age, training received on rice production, agricultural extension media contact and problems faced in practicing rice yield gap minimizing strategies had also significant contributing factors at (p < 0.05) 5% level of significance on boro rice yield gap minimizing strategies practiced by the farmers.

In summary, the model suggests that the respective authority should consider farmers age, educational qualification, yield gap of boro rice, training received on rice cultivation, agricultural extension media contact and problems faced in practicing boro rice yield gap minimizing strategies when made policy for their extent of adaptation strategies towards salinity effects in agriculture to be improved.

Data furnished from Table 5 , revealed that, farmers age had a positive influence on their practice of boro rice yield gap minimizing strategies. Data also showed that farmers age had a significant contributing factor (p < 0.05) on their practice of boro rice yield gap minimizing strategies. It could be said that middle and old aged farmers were given more preference to rice cultivation than the young aged. It might because the old aged farmers had land ownership, comparative experience in farming activities and were more involved in rice cultivation than the young and middle aged farmers.

Data revealed from Table 5 , showed that, farmers educational qualification had positive influence on their practice of boro rice yield gap minimizing strategies. Farmers educational qualification had the most significant (p < 0.01) contributing factor on their practice of boro rice yield gap minimizing strategies. It seemed that educated farmers had more knowledge and greater ability to understand and respond to anticipated changes, had greater access to information and opportunities than others, which might drive them to adopt advanced strategies to minimize rice yield gap.

Data revealed from Table 5 , showed that, yield gap of boro rice had a negative influence on farmers practice of boro rice yield gap minimizing strategies. Data also showed that yield gap of boro rice had the most significant (p < 0.01) contributing factor on farmers practice of boro rice yield gap minimizing strategies. It was found that farmers practicing medium and maximum level of strategies had low yield gap in the study area. In the study area it seemed that low yield gap inspired farmers to follow more advance strategies to get higher yield in future.

Data furnished from Table 5 , showed that, training received on rice cultivation had positive influence on farmers practice of boro rice yield gap minimizing strategies. Training received on rice cultivation had significant contributing factor (p < 0.05) on boro rice yield gap minimizing strategies practiced by the farmers. It seemed that when farmers receive different types of training their knowledge and skill increase, they become more aware of adopting different boro rice yield gap minimizing strategies to get higher yield.

Data revealed from Table 5 , showed that, agricultural extension media contact had a positive influence on farmers practice of boro rice yield gap minimizing strategies. Agricultural extension media contact had a significant contributing factor (p < 0.05) on farmers practice of boro rice yield gap minimizing strategies. It seemed that agricultural extension media contact helps farmers to get proper and advance techniques of rice cultivation, the guidance of production activities, provide up to date advice on what to do in every stage of rice cultivation and forecast in outbreak of diseases, pests and in natural calamities. Due to those factors farmers having well agriculture extension media contact might have low yield gap.

Data furnished from Table 5 , showed that, problem faced in practicing rice yield gap minimizing strategies had negative influence on farmers practice of rice yield gap minimizing strategies. Problem faced in practicing rice yield gap minimizing strategies had significant (p < 0.05) contributing factor on farmers practice of boro rice yield gap minimizing strategies. In the present study, it was found that farmers facing less problems in rice production got higher yield than the farmers facing more problem in rice production. It may because, low problem in rice cultivation enhances higher production and minimize yield gap.