proceeded Dr. Onen as facilitator covering the sub topic "Basic Research Approaches and Designs: An Overview" with his presentations for the third day running. In first shift of the morning the facilitator covered the definition of Research Design, meaning of Research Methodology, Criteria for Classifying Research Designs, Research Designs/Approaches by Research Paradigms, Specific quantitative research designs, Specific Experimental Designs, Elements of the Research Design, Example of Design Notations, Non-Experimental Quantitative Designs, Qualitative Research Designs, Types of Mixed Methods Research Designs including Sequential Explanatory Design, Sequential Exploratory Design (SED) and Concurrent Triangulation Design (CTD) and Questions to answer when choosing research designs.
In the second shift of the morning session Dr. Onen covered Study Population and Sampling, meaning of a Study Population, definition of the Accessible Population, how to define a Sample, and what Sampling is, and further the reasons for Sampling, in addition to weaknesses of sampling. Other items covered were the Common Terms Used in Sampling such as Randomization, Representativeness, Parameter, Statistic and Precision. Furthermore other areas covered included Sampling Strategies, Probability and non probability sampling, Summary of Common Sampling Techniques, Sampling Techniques such as Simple Random Sampling (SRS), Stratified Sampling, Systematic sampling, Cluster Sampling, Snow-ball or Networking, Purposive Sampling, Convenience Sampling and Quota Sampling. He opined that the choice of a sampling technique is guided by the objectives/ hypotheses/ questions of the study. The facilitator further taught about the purpose of the sampling technique in which he emphasized that each sampling technique is intended to achieve a particular purpose and that for the research to be meaningful the purpose of the study must tally with the purpose of the sampling technique.
He urged careful examination of the relationships between the variables in the study. On Techniques of Determining Sample Size, Dr. Onen stated that there exist two approaches the first being Non-Statistical Estimations e.g. Kathuri and Palls (1993) propose a conventional rule of thumb that for survey each major group should have 100, each minor group 20-50; in ex-post facto/ experimental - 15 in each condition and correlation - 30. However, the larger the sample the better the sample. The second approach is use of Statistical Methods where we researchers could either determine Sample Size through Power Analysis or use of Statistical Table. Lastly he also mentioned the method of adapting Sample Size from studies whose credibility is proven, however he cautioned that one must Avoid Sampling Error and Sampling Bias. The last item covered was Sampling in Qualitative Studies in which Dr. Onen indicated that there are no rigid rules preferred yet, however, most significantly, the common view currently rests on the researcher reaching the saturation point with responses from his respondents. He also explained in great detail how to write the sampling techniques in research work.
The Questions and Answers (Q&A) session yielded a few questions which were given responses as noted in the proceedings.
A member which to know how a researcher would determine accessible population, or is there is any existing formula to achieve that. Dr. Onen answered in the negative and there is really no formula but a researcher can restrict and direct himself through the design and variables as defined in the work to determine accessible population. For example by restricting oneself to two out of ten universities to represent views of all the other universities in Somalliland already one is defining one's accessible population. Another question arose on how a researcher can prove whether the confidence level in a research work is suitable for that study or not. In his response Dr. Onen stated that authorities and tradition in various fields of specialization will almost always likely guide one on determining confidence level. One member wished to know if in a case where the population is too small, there is any need to use Krecji and Morgan table to determine sample size. According to Dr. Onen perhaps this may not be really very suitable. It would be better rather to use the whole population for very low numbers like 15 to 30 while drawing from the philosophy that the bigger the sample the better.
Another member inquired on what risks are associated with use of sample size from previous studies. In his response he said that if the study does not fit properly in terms of similarities and if everything is not very well contextualized, results may not necessarily yield accurate outcomes, hence there is need for extra caution if one was to go that way. A question was posed as to whether one is allowed to use higher figures than what a table has specified if one decides to use a table. Dr. Onen's answer was a strong no. He defended this answer by saying if one wishes to use a different figure from the one offered on a table, one should apply an appropriate formula, in order to determine one's sample size rather than go for teh table and decide not to take the correct value so that one can be as scientific as possible.
There was an extended debate on adoption of qualitative research approach versus quantitative research approach for various studies. In this regard Dr. Onen enquired whether Amoud University has had any research done using qualitative approach and it was realized that since establishment of the postgraduate programmes, not even one thesis has been submitted with qualitative approach. He wondered loudly why and advised strongly that students and supervisors in areas where qualitative research is applicable must be equipped, and very strongly encouraged to explore that option because it would be a disservice to other fields of specialization to arm twist all work to go quantitative even where the situation calls for a qualitative approach. However, he emphasized that it would be very important to appreciate the fact the qualitative research majorly focuses on richness of information and the skills plus tools of data collection rather than predefined sample sizes, which often makes the sample size fairly smaller compared to quantitative approach. Students must be allowed to go either qualitative studies and quantitative studies, he stressed.
A member asked to the difference between descriptive studies and explanatory studies. Dr. Onen explained that the major difference is that descriptive studies are applied to both quantitative approaches and qualitative approaches whereas explanatory studies are mostly qualitative in nature. The last question was on reliability and validity where a member wished to know whether Amoud University could use its own formula in establishing reliability and validity. Dr. Onen gave a strongly and solid objection emphasizing that formulas for calculating must such important aspects in research as reliability and validity should be carefully chosen especially from tried and tested authorities in the field of statistics. However, if an institution or individual wishes to replace such an important formula it must presented to the academic arena to be interrogated and proved by peers in reputable journals and other academia platforms before being adopted so as to avoid the risk of academic fraud.