Confidence/Error & Parsimony 8/8 Characteristics vs Hallmarks خاصية/ميزة معامل الثقة والخطأ،والتبيسط — Transcript

Full transcript

1. 0:00welcome your honors
2. 0:02this is the last episode which is about the last two terms precision and confidence
3. 0:12error and parsimony
4. 0:16in the end, we will comment on which of these are characteristics and hallmarks
5. 0:20precision and confidence
6. 0:22In fact, precision here does not function as a term or a concept more than it is a meaning of two components
7. 0:35one is confidence and the other one is error
8. 0:39so what are confidence and error?
9. 0:40let's give a very simple example
10. 0:42let's suppose that you are measuring phenomena like commitment, which we discussed last episodes.
11. 0:49and you are measuring it with previous knowledge of a number of variables
12. 0:55I will mention a metaphor here, you want to cook something specific, for instance, a cake
13. 1:06we made a cake which means we made a commitment
14. 1:11commitment is the phenomena, a cake is the phenomena
15. 1:14and we studied the components which are the problem and literature
16. 1:17then we apply it
17. 1:19so what happened in the example of the cake?
18. 1:22the cake was eaten and those who ate it are asking you to ask them for their opinion
19. 1:29and you as the research you are willing to know
20. 1:32you found that people who ate the cake have two different opinions they make two groups 50/50
21. 1:36one group says thank you for the cake
22. 1:41while the other group says thank you for something totally different pizza for example
23. 1:47as if you are working on commitment but when I applied it, it has been achieved on only 50% of respondents
24. 1:57and the other 50% they showed no result for commitment
25. 2:01and this category understood commitment as something absolutely different
26. 2:04the activities which are an application of outcomes, findings, and recommendations from my research
27. 2:10applying those variables people understand as loyalties
28. 2:1450% understood it as something absolutely different
29. 2:23this is a problem, and as you are the researcher you understood the problem
30. 2:30and you are thinking about a solution to it
31. 2:31then you have decided to conduct another research
32. 2:33this time I will do something different, I have studied the variables in a different way
33. 2:41and I brought its components, I studied that then I interviewed people
34. 2:46what did I do here? let's get back to our example
35. 2:49you brought a new cake, then invited people to eat it
36. 2:53they ate the category we are studying and applying for it commitment
37. 2:59this time, everyone 100% thanked you for the cake, not like the first research where it was 50/50
38. 3:112 or 3 of this group, let's suppose that the group is formed of 100 individual, they thanked you for the cake
39. 3:20but they have some remarks
40. 3:24the first one said it is sugary, the second one said it is salty, the third one said it has a flavor I did not like
41. 3:32this situation has a different name, it is not like the first case
42. 3:34the first case all the people took two different stances
43. 3:39either you brought the cake or not, they considered it as something totally different
44. 3:44it is called confidence, confidence here as if it is not a valid measure
45. 3:50the validity of the measure here has a problem
46. 3:53therefore, you can slightly relate confidence with validity
47. 3:57did you measure what you were supposed to measure?
48. 4:01this is the closest meaning to it, did you measure what you were supposed to measure?
49. 4:05then, you have a problem here
50. 4:07both categories understood that you measured and applied something different
51. 4:12those say cake and those say not a cake
52. 4:17those say that this is a commitment that you were studying and the others say that is loyalty
53. 4:22then here there is a problem
54. 4:24then what is the name of the first case? it is called error
55. 4:28why? because the majority of people even if it is not 100%
56. 4:32let's suppose it is 100% said that is a cake
57. 4:36then what is the issue? it is solved. No, this cake is not tasty for them, one like the cake with less of sugar
58. 4:42and another one with slightly more sugar, those people having opposed tastes
59. 4:46as it is obvious these two do not belong to the nine elements or our context
60. 4:49it was necessary two do one of these two
61. 4:52either to define the context in the group that will like the cake
62. 4:56and this could be a mistake in defining the problem, which could be the reason for a sequence of mistakes
63. 5:04in the steps of the research or there were variables in data
64. 5:08that you should have collected in the review of literature, research design, and the data collection
65. 5:13which could have been suitable for the needs of these people
66. 5:19secondly, the first category where were two groups are called confidence
67. 5:25so, to what extent this confidence is good, we will talk about it
68. 5:30and in the second case, where there is a great confidence 90, 99, 100% but still there is a small group
69. 5:382, 3 or 4 who did not like the cake but they admit that it is a cake
70. 5:45here there is a problem of reliability, not validity as if you are making a measure but each time you measure it you find it different measures
71. 5:54as if it is tolerance
72. 5:55this is called the error
73. 5:57in this form, if we want to talk about confidence and the error that is both
74. 6:04confidence and error they form a total of precision
75. 6:09we advise of certain numbers in the confidence and here I am talking about the domain of business
76. 6:18that the minimum of the confidence must be no less than 75-80%
77. 6:23only if I reviewed the literature and it indicates that even if this is about business and it mustn't be under 90%
78. 6:30in this case, we will obey to the rule
79. 6:32generally, please pay attention to confidence, and tell us how much confidence you targeted
80. 6:42this we will use it in two bigger areas, or three. The first area is the literature review
81. 6:50you have chosen a model that achieves the confidence
82. 6:54am I making a model that can help me achieve my results with of percentage of 75% to 80% while measuring commitment or these phenomena?
83. 7:03yes, it will.
84. 7:05in the research design, I have materials that will preserve this 75-80% reduce or increase it
85. 7:14and where is the third component? when you collect data
86. 7:19and always pay attention that sum of this two confidence and error is not 100%
87. 7:29because the sum of each one is 100%
88. 7:34when I move to the step of testing, this is the third step where you are going to mention the six steps
89. 7:40confidence and error mentioned one time because of your choice from the literature quality of the science or the model
90. 7:47the second time in the research design. how did your research design respected, preserved, increased, or decreased confidence and error?
91. 7:56and in the third step when you do a test, we want you to do a test and this is the materials we need in the test
92. 8:03Cronbach alpha, chi-square, t-test and f-test materials of examination, using them depending on our need and research
93. 8:12tell us if you found the same number or not
94. 8:15in error we want to target no more than 15-20%
95. 8:23and sometimes we can substitute the error by p-value
96. 8:27it is possible if we have studied physical sciences such as pharmacy, medicine, engineering or other
97. 8:36we could have noticed this term in the field of medicines or in the patient information leaflet
98. 8:42in some scientific research, mathematics or correlations, statistical analysis and so on
99. 8:48this p-value that indicates the error mustn't be more than 15-20%
100. 8:53one more time, I am talking in the field of business. It could be in medicine or engineering
101. 8:59in any critical field which affects the human health and the value-added and applied or impact, we will increase that
102. 9:06even in the business, if we are talking about billions or in the common good or whatsoever
103. 9:10we will increase that tI we reach 90-95 and possible to 99
104. 9:13is that doable? yes, but I will always have the challenge in the error
105. 9:17as long as we minimize the error it will be difficult for us
106. 9:21because we have a valid measure yet it is not fully reliable there is chances for mistakes to happen
107. 9:27this absolute error is very difficult yet possible and it is the cause of special scientific results
108. 9:35if we continued this way the researcher has two options
109. 9:39when he/she finishes the research design the researcher will collect data analysis
110. 9:43conclusion, there are two main options, either he/she will achieve exactly what mentioned in the research design
111. 9:51this way he fulfilled the aim, he achieved the minimum
112. 9:54or the researcher could not achieve it although he/she has fewer numbers 30 or 60%
113. 10:0160 is less than this 75-80%
114. 10:03and this 30% is not suitable because it is bigger than the error percentages allowed
115. 10:09then this marked in black can not be accepted
116. 10:13then what will those who are better than the normal do?
117. 10:17they can either fix the research design or collect strong data which will make us achieve a higher level of confidence and validity
118. 10:2675-80% to 90%
119. 10:30and succeeded in increasing the average of error which means he/she enhanced the reliability
120. 10:35through minimizing the error to 5% rather than the targeted to 15%-20%
121. 10:39this means that the researcher added value
122. 10:42and here I am calling you to think, are the confidence and the error are characteristics or hallmark?
123. 10:52mandatory or added value?
124. 10:54let"s discuss the last one parsimony, parsimony means and you have it in the majority of books
125. 11:00in Arabic means greed and it has many synonyms
126. 11:06the targeted meaning here is simplicity
127. 11:10suppose that you are in the review of literature, these two are the one which is substituted by precision
128. 11:22if you are reading the literature and you found a model A and three variables and these model is promising to measure
129. 11:37the phenomena for instance commitment 80% confidence and 15% error
130. 11:42and you found another model B, in green, it has many variables 1.2.3.4.5.6.7
131. 11:49measure the same phenomena but it is promising you with 90% confidence and 5% error
132. 11:55and I am conducting scientific research, which one am I going to choose?
133. 11:58it is normal to choose B, why? because it is the strongest
134. 12:03what are the reasons for it? because it has more numbers of variables
135. 12:07and pushes us to believe that it can achieve, measure and collect all the variables
136. 12:12which are better affecting the phenomena
137. 12:14and what is more important for us is that it measured the ability of the model as validity with 90% confidence
138. 12:25and the supposed error percentage and measured for this research in a scientific way average of error is 5%
139. 12:30but this is less than 15-20%
140. 12:32but it has a high quality, it is as we are saying here that we do not have any problem in choosing model B
141. 12:39but you have a condition which is parsimony simplicity
142. 12:44you must include both of them in the literature and we will discuss this with detail in the literature review
143. 12:50but to apply the principle of parsimony I will choose A
144. 12:57why? because we are telling you that your choice to model B is a burden
145. 13:03because it does not include only variables and this
146. 13:06there will be some conditions as a number of data
147. 13:09specific time for collecting data, specific method in dealing with variables, respondents
148. 13:14environment and so on
149. 13:16use the simple one with a condition, of course, to not be less than what was agreed on in science
150. 13:24science of this domain if it is business, medicine or engineering
151. 13:28and see what science approves here
152. 13:31in general, we try, in the business management even in the social studies, it should not be less than 75-80%
153. 13:40and do not exceed 15-20%
154. 13:43here 80% confidence and 15% error is fair, you can used
155. 13:46and you will have also your own variables which you added to the problem
156. 13:52as of 8-9-10 if we added them here you will have 6 variables and here you will have 10 variables
157. 14:03your ability as a research to study, apply and examine model from 6 variables 1 to 3 plus the other 3 that are specific to your environment
158. 14:11plus 10 variables it will be easier and simpler here
159. 14:16we want strong simplicity
160. 14:19we do want complication and then the research will be weak
161. 14:22and this way we will have an issue
162. 14:24is the principle of parsimony a characteristic or a hallmark?
163. 14:28let’s revise them at the end of this episode which well try to integrate
164. 14:33all the previous episodes characteristic or hallmark
165. 14:36is purposiveness a characteristic or hallmark?
166. 14:43possessiveness is supposed to be a characteristic
167. 14:47what is a characteristic? means mandatory and what is mandatory?
168. 14:49means it is a must for the research
169. 14:51because it means that there are variables by which we will measure the phenomena
170. 14:57which we are studying
171. 14:58the absence of variables makes it almost impossible to measure the phenomena
172. 15:06and here in relation to social studies and physical studies or science
173. 15:11and rigor is it characteristic or a hallmark?
174. 15:17what is rigor?
175. 15:19means a theoretical base and a good methodological design
176. 15:26that we agreed on is a literature review
177. 15:28and research design
178. 15:30if I have the possibility of excluding the steps of review of the literature and research design
179. 15:34this way I will be ruining deductive approach and I will not know how to have an applied research
180. 15:40It is mandatory, it is a characteristic
181. 15:44testability, what is it?
182. 15:48there is a difference between a test and testability
183. 15:51the place of the test is in the data or analysis
184. 15:54but testability means that I have a hypothetical statements
185. 15:59so what is the basis of these hypothetical statements? review of literature
186. 16:05integrated with the problem
187. 16:07if there is no hypothetical statement then I do not have a sentence
188. 16:12that indicates variable the relationship and the phenomena in the part of research design
189. 16:15then what does data mean? this way we have a problem
190. 16:18the testability; therefore, means that you have a hypothetical statement is a must
191. 16:23and pay attention here, it is considered a must in applied research which is deductive
192. 16:30why? one of its names is Hypothetico deductive
193. 16:35it is responsible for examining hypothetical statement supported by science
194. 16:40therefore, testability is a must, mandatory
195. 16:45the episodes were in order, after that we had objectivity, not this serial order
196. 16:50objectivity, slightly easy. mandatory
197. 16:54objectivity, is a mandatory characteristic why?
198. 16:57because all the components of scientific research steps, sub-components, and the items must be integrated
199. 17:04objectivity, slightly easy. mandatory
200. 17:09replicability means that my research is subject to be repeated
201. 17:16in similar circumstances, we can not say that it is mandatory
202. 17:23it is a hallmark
203. 17:24and hallmark means it distinguishes and adds value
204. 17:29but not mandatory for the researcher
205. 17:31because you have done research that fulfills the condition of the environment,literature
206. 17:36and design, collected data correctly
207. 17:38analysis conclusion is fair enough
208. 17:41to be accepted for your degree
209. 17:43or the scientific research you are publishing
210. 17:45you are not supposed to do more
211. 17:47it you achieved replicability it will be an added value
212. 17:50replicability could be the first element here to consider as a hallmark
213. 17:55generalizability goes without saying if the replicability was which has a similar methodology or research design and outcome
214. 18:12then if I have different steps I am not obliged to do it
215. 18:15my research is applicable in another environment
216. 18:20or different
217. 18:22it is not easy or acceptable
218. 18:24then, generalizability is in the first place a hallmark
219. 18:29precision which indicates to confidence and error
220. 18:34is it a characteristic or hallmark?
221. 18:46in fact, it could be the two
222. 18:49how? if you achieved the minimum 75-80% confidence and the maximum 15% error
223. 18:56you adhered to from the literature
224. 18:59and from your development to the research design
225. 19:03and you adhered to them in the test and gave us results
226. 19:05matching that, and you prove validity and reliability that they are the same you targeted
227. 19:11then you fulfilled a mandatory characteristic, minimum
228. 19:17but you are special research either you chose a difficult research design or literature
229. 19:26but here you achieved 90% confidence and 5% error
230. 19:30then you have given us something special
231. 19:31you gave us an added value
232. 19:33let 's say that this research is of higher quality in a standard research
233. 19:39then we can consider the error and confidence a mandatory characteristics
234. 19:44and we can consider it an added value if you adhered to the minimum value
235. 19:48or the maximum and you achieved more than it in the analysis and in the outcome
236. 19:53parsimony, in fact, it is
237. 19:57I can choose at the beginning the B
238. 20:01what will this B do then?
239. 20:03I am afraid and I feel embarrassed complicated, strong, comprehensive model
240. 20:12that can achieve 90% confidence and 5% error and I will not be able to have results
241. 20:16or even the results become bellow 80% and 15-20%
242. 20:22then I will choose the safer track
243. 20:28applying the principle of parsimony is a minimum characteristic or the safest solution
244. 20:38because of choosing this model
245. 20:40with a condition, it should not be less than 75-80% and 15-20%
246. 20:45I decided not to apply parsimony
247. 20:52and you will go for the hard one
248. 20:56then you have given us an added value a hallmark
249. 20:58in brief, to conclude the episode that discussed hallmarks and characteristics
250. 21:07purposiveness is a characteristic
251. 21:09rigor a characteristic
252. 21:11testability a characteristic
253. 21:13objectivity a characteristic
254. 21:15replicability and generalizability are hallmarks added value not mandatory if you could do them then they will be a hallmark
255. 21:26precision which stands for confidence and error can characteristics if you adhered to the minimum if you go for the maximum than that of the science
256. 21:39they are hallmark
257. 21:40as well as parsimony, using it will be a characteristic in the simple version
258. 21:46and using or not using it in the complicated model then it is a hallmark
259. 21:58Thank you, your honors by this we have concluded the episodes of characteristics and hallmarks
260. 22:04the next episode will discuss a new topic
261. 22:09we will talk about forms of data and information
262. 22:13and we discuss the beginning of literature review
263. 22:15thank you