1. Do start from an observation that is related to a business behavior and that may lend itself into a signficant business related phenomenon through gures and statistics. Do validate a number of research questions related to the phenomenon through surveys and discussions with SMEs who should validate the subject matter of the phenomenon and the related questions that are of business concern from their point of view. We frequently notice that the descriptive analysis results which students state in their papers and present during the defense do not really promote a problem, rather, they conrm a healthy status; customers are satised, employees are happy and engaged, projects are successful, performance is good, etc. This happens because SMEs were unable to see the targeted issue during the initial interviews for various reasons. To overcome this initial hurdle, students are advised to conduct other rounds of surveys with other stakeholders such as clients, partners, suppliers, regulators, etc. In other words, you are asked to do a research before you do the research. Don’t worry! Should your problem be proven valid, this rst research will form the major components of chapter 1 and 2 of your project. At this stage you will be looking at quantitative data from statistical reports to prove there is a pattern or a phenomenon, and you will be conducting a formal qualitative research by interviewing the SMEs using proper scientific methods to establish the significance of the phenomenon. By scientific methods we mean designing the interview script and session, determining the criteria for the selection of stakeholder and performing qualitative data analysis. Many students wrongly believe that this preliminary qualitative research is quick and informal. You should be alerted to the fact that the quality of your subsequent phases of your research is dictated by the quality of this preliminary qualitative research during the \Phase Zero". 2. Don’t stay at the macro level of the topic and the subject matter. Broad subjects and general approaches are unacceptable, for they don’t have a value nor can they invite properly focused research questions. Do narrow down your topic and aim towards the specics of the subject matter where the value is. 3The \Phase Zero" is a term we have developed based on the observations we had record during the defense session. By \Phase Zero" we refer to the initial research students must perform in order to complete an approved proposal. Should students perform a proper \Phase Zero", they will inevitably be able to lead a proper selection of the topic and a valid research problem. The \Phase Zero" consists of 7 steps; Observation, Verication, Validation, Literature Review, Signicance Assurance, Research Problem Denition and topic, Research Methodology. 4 1. Project Topic Selection and Problem Validation Issues the \Factors aecting . . . ", \Reasons lead to . . . " or \Obstacles/Challenges that hinder . . . ". These questions are either already answered or can be easily answered through means less of a formal scientic research. 3. Do, during the \Phase Zero", a proper literature review and proper interviews with the SMEs to avoid searching for the already found. Do in-depth reading to explore the relevant theories and models in order to nd areas of new interesting or dierent perspectives that would lead you to new interesting and specic research questions that are of added value to concerned stakeholders. Do focus on few factors/reasons or a specic factor, construct such as trust. In other words, this specic factor or reason could act as a new dependent variable (Y) for which you need to think of another set of independent variables (X’s). Doing so, you may nd something was not discussed before adequately. You can continue digging deeper and deeper till you nd interesting and researchable level where the research question promises new value added. 4. Don’t do a research question that is easy to answer and easy to collect data for, for your research will be of no value specially when the answer is already known to the SMEs or the concerned managers. In the case of B2C topics for example, don’t collect your data from customers only. Always Do look for topics that have some challenges in accessing the data. Challenges do not mean condentiality, it means the data was not known before your research and are still in the mind of the senior executives that are hard to reach, as an example. 5. Do identify all data contributors (360 degree research) and select the most signif- icant data holders. Your selection of the dierent stakeholders should be based on the variant valuable views of each, see point 1 above. 6. Don’t research the \Intention". Do research the \Action". If you want to know why employees leave their company, don’t ask the current employees about their intentions to leave, you must seek those who have already left, the ones who still have fresh experience. 7. Do crash your Research Question. You may need to expose your research question to harsh conditions, in other words, perform a crash test. Should your research question survive the test, you can be assured that you have secured a well matured research question. The following 7 Steps are the needed procedures to perform the crash test: a) Dene the Y, the dependent variable. Take your time to do a very careful denition. b) Dene the X’s, the independent variable(s) (your bias, your emotion, your claim). X is the solution or the victim that you blame for why Y is changing and behaving unexpectedly. X is your claim that you think will solve the problem/issue of Y. X’s could be the technology, the strategy, or the policy 5 2. Literature Review Issues that you will recommend at the end of the project. If the X’s are known(based on SME’s discussion or literature review), then change the X’s, go deeper. c) Think about the unit of analysis, company, business unit, employee, customer, time, product, etc. Think about a spreadsheet structure that has X’s and Y as columns and think about what could be possibly the rows of the spreadsheet; usually they are the unit of analysis. If the X’s and Y are not harmonized over the same unit of analysis (e.g. Y is about companies, and X’s are about employees), then you may need to change either X’s or Y. d) Check whether X’s & Y are categorical or numerical data type. It is acceptable to have dierent types of data, however each data type will require a dierent tool for the data analysis. e) Check whether both X & Y are changing or not. If both are changing then, you are on the right track, otherwise you are not. Then this means you are doing your research on a variable that does not change which will not create any relationship or eect with the other variable. f) Think about the Z variables as columns in the Excel sheet where you may want to apply the lter option. Imagine, you lter your dataset using the Z variables by selecting certain values such as sector, department, gender, product, etc. You should get subset of a data that you need to apply your data analysis tools to. By doing so, you will be able to discuss the generalization, the limitations, and the validity of the ndings. g) Check in the spreadsheet (see above) how many rows you ended up with. If they are small, then probably you may need to think consider a qualitative research otherwise go quantitative to nd statistical signicance.