Botswana’s livestock sector accounts for about 70 percent of foreign exchange earned from the agriculture sector. Livestock is a major source of income, employment and provides a store of wealth to many low and unskilled Batswana. However, the livestock sector is threatened by several risks including pests and diseases, climatic shocks and abrupt policy shifts that affect production. Some of these risks can be mitigated by adopting livestock insurance. In 2010, the Botswana Insurance Company (BIC) introduced a livestock insurance policy with the objective of encouraging farmers to take up measures that reduce the adverse effects of natural calamities. However, despite the existence of livestock insurance policy, few traditional livestock keepers have adopted it. The objective of this study, therefore, was to assess smallholder cattle keepers’ willingness-to-pay (WTP) for cattle insurance attributes (as a non-market commodity) in the Central District, the largest traditional cattle-keeping district in Botswana. Five focus group discussions (FGDs) were conducted in Mahalapye, Radisele, Tamasane, Palapye and Serowe villages to determine the main attributes that the livestock keepers would want to have in a cattle insurance policy. Using the factorial design in conjoint analysis, six attributes and their levels were combined into sixteen profiles which were presented to respondents in pictorial format for preference ranking. The profiles were presented to 182 random selected smallholder cattle keepers for their expression of WTP. The WTP was derived from the marginal utilities of insurance attributes. An ordinary least squares regression was used to assess factors influencing the WTP for cattle insurance attributes among livestock keepers in the study. The results show that the livestock farmers were willing to pay $17.77, $7.07 and $1.12 for weather index-based insurance cover, covering proportion of a herd and replacing a dead cow with a live one respectively. The overall xmean WTP for a full insurance product was $11.45, which is within the price range of the current livestock policy used by commercial livestock keepers in Botswana. The factors that positively influenced respondents’ WTP for cattle insurance were distance to the nearest tarmac road, off-farm investment income, total land size owned by the respondent, and total livestock unit (TLU). The age of household head, access to credit, annual crop sales income and vaccine cost negatively influenced the WTP. In addition, having extra income, land and many cattle increased livestock keepers’ WTP for cattle insurance. Being an older farmer who vaccinated his livestock and received cash from arable farming reduced the WTP for the cattle insurance policy. Given that the most preferred cattle insurance profiles were numbers 7, 12, 10 and 14 with WTP values of US$23.88, US$20.13, US$18.88 and US$18.13 per unit respectively, the BIC could consider designing cattle insurance policy products within that price range. The most ideal policy should have the attributes of profile number 7, i.e., have a 1-month compensation period, cover a portion of the cattle herd, be weather index-based at a monthly premium of US$0.7 paid as an annuity, and compensate keepers with a live animal in case of a loss. Keywords: Willingness to-pay, Cattle insurance attributes, Preference, Smallholder livestock keepers, Conjoint analysis, Ordered probit model, Central District, Botswana

The purpose of this study was to establish whether Computer Adaptive Testing can be successfully adopted in educational assessment in Kenya by focusing on factors likely to affect its adoption. The study objectives were to determine a) the extent to which adoption of CAT is affected by students and teachers perceptions b) the extent to which adoption of CAT is affected by the existing technological infrastructure; and c) the extent to which adoption of CAT is affected by levels of computer literacy. The study used a sample selected from public secondary schools in Makueni County through simple random sampling technique. Questionnaires with both open and closed questions were used to collect primary data from the respondents while secondary data was obtained from the Ministry of Education, Makueni County Office. The data was analyzed using the Statistical Package for Social Sciences (SPSS) where Descriptive statistics such as frequency distributions, percentages, means and standard deviations as well as inferential statistics such as Pearson Correlation and ANOVA tests were utilized. The findings of the study indicate that teachers and students perceptions are positively related with adoption of CAT (0.085) although the relationship is weak; there is significant relationship between ICT infrastructure and adoption of CAT (0.336) and the relationship between teachers and students level of computer literacy and adoption of CAT is negligible (0.049).