Methodological issues are one of the prime considerations for conducting research for yielding valid findings. Methods and procedures that were followed in this study are discussed. Multi-stage random sampling procedure was followed in this study. At first Sadar upazilas of Dinajpur, Thakurgaon and Panchagarh district were purposively selected due to climate vulnerable (drought prone) condition of the farmers of the study area. Secondly, two unions from each sadar upazilas were randomly selected followed by two villages from each union. Therefore, 12 villages of three districts were the ultimate locale of the study. Lists of 960 farmers from 12 villages were prepared with the help of the Sub-Assistant Agricultural Officer (SAAO) and Agricultural Extension Officer (AEO) who are mostly affected by drought and vulnerable to climatic conditions. About 10 percent farmers from 960 population i.e. 96 farmers were selected by using simple random sampling procedure as sample of the study. A reserve list of 12 farmers (about 10 percent of the sample) was prepared so that these farmers could be used for interview in case of unavailability of any farmer included in the original sample in spite of utmost effort during collection of data.
The selected characteristics of the farmers were age, educational qualification, family size, farm size, annual income, training received, credit received, climate change knowledge and extension media contact. Proper statistical techniques and procedures were followed during measurement of these characteristics. Severity of climatic hazard was measured by 4-point rating scale such as not at all?, mild?, severe? and „most severe? and weights assigned to these responses were 0, 1, 2 and 3, respectively. The farmers were requested to indicate their response with each of the selected climatic hazard. Severity Score (SS) for each five climatic hazard was computed using the following formula. This formula is prepared by modifying the formula of Utilization Score (US) used by Haque et al. (2019).
SS= (Pnaa×0) + (Pm×1) + (Ps ×2) + (Pms×3)
Where, SS= Severity Score, Pnaa = Percentage of the respondents indicating climatic hazards as not at all?, Pm = Percentage of the respondents indicating climatic hazards as mild?, Ps = Percentage of the respondents indicating climatic hazards as severe?, Pms = Percentage of the respondents indicating climatic hazards as most severe?
Severity Score (SS) for any one of the selected climatic hazards could range from 0 to 300 where 0 indicated no climatic hazard and 300 indicated most severe.
The use of information sources was the focus issue of this study. To measure the use of information sources of a respondent 17 information sources were identified during pre-testing of the interview schedule. To measure the use of information sources three dimensions namely; (i) Availability of information sources (ii) Reliability of information sources and iii) Appropriateness of information sources used by the farmer for climate change adaptation were studied. A similar dimension was used by Mwalukasa (2013) in his study on agricultural information sources used for climate change adaptation in Tanzania. Availability of information sources of the farmer for climate change adaptation was measured by 4-point rating scale such as not at all? low available?, moderately available?, and highly available? and weights assigned to these responses were 0, 1, 2 and 3, respectively. The farmers were requested to indicate their response with each of the selected information sources. Thus, a respondent's information sources availability score was obtained by adding the weights for their responses to all the 17 information sources incorporated in the questionnaire and hence, availability of information sources score could range from 0 to 51, where 0 indicated not at all available and 51 indicated highly available with the 17 information sources. Reliability of information sources of the farmer for climate change adaptation was measured by 4-point rating scale such as not reliable? fairly reliable?, reliable? and very reliable? and weights assigned to these responses were 0, 1, 2 and 3, respectively. Appropriateness of information sources of the farmer for climate change adaptation was also measured by 4-point rating scale such as not appropriate? fairly appropriate?, “appropriate?, and very appropriate? and weights assigned to these responses were 0, 1, 2 and 3, respectively. Similar scaling technique was used by Mwalukasa (2013). For clear understanding and comparative analysis of information sources index was calculated. Information Availability Index (IAI) for each 17 information sources was computed using the following formula:
IAI= Pnaa×0 + Pla×1 +Pma×2 +Pha×3
Where, IAI=Information Availability Index, Pnaa= Percentage of the respondents indicating information sources as Not at all?, Pla= Percentage of the respondents indicating information sources as Low available?, Pma= Percentage of the respondents indicating information sources as Moderately available?, Pha = Percentage of the respondents indicating information sources as Highly available?
Information Availability Index (IAI) for any one of the selected sources could range from 0 to 300 where 0 indicated information sources not available and 300 indicated highly available. Information Reliability Index (IRI) and Information Appropriateness Index (IApI) was also calculated similarly as Information Availability Index (IAI)
An interview schedule was prepared mostly with close-form questions. Simple questions and statements were included in the schedule to obtain information considering the research topic and objectives. In the survey, data were collected through face to face survey by the researcher himself. Data were collected during15 April to 15 May 2017. Data obtained from the respondents were coded, compiled, tabulated and analyzed in accordance with the objectives of the study. Qualitative data were converted to quantitative data by means of suitable scoring to facilitate analysis and interpretation. The analysis was performed using Statistical Package for Social Science (SPSS) 22.0 computer package. Descriptive analysis such as range, number, percentage, mean, standard deviation and rank order were used whenever necessary. Pearson's Product Moment Correlation Coefficient (r) was used to examine the relationships between focus issue and selected characteristics?. At least a 5 percent (P=0.05) level of probability was used as a basis for the rejection of the null hypotheses.