WEBVTT 00:00:00.630 --> 00:00:05.060 Grounded theory is a common approach for doing qualitative research. 00:00:05.060 --> 00:00:08.480 The Gioia method is one implementation of this approach. 00:00:08.480 --> 00:00:14.220 The term grounded theory can be understood in two different meanings. 00:00:14.220 --> 00:00:18.669 In a broad sense, it can be understood as any approach, 00:00:18.669 --> 00:00:21.009 where theory comes from the data, 00:00:21.009 --> 00:00:24.630 so that you don't have any prior theory that you're testing. 00:00:24.630 --> 00:00:27.210 Instead, you're analyzing your data, 00:00:27.210 --> 00:00:30.029 and then the theory emerges based on that data, 00:00:30.029 --> 00:00:33.380 or you get new ideas about a theory based on the data. 00:00:33.380 --> 00:00:35.460 So that's what it is about. 00:00:35.460 --> 00:00:38.340 Another way of understanding grounded theory is that, 00:00:38.340 --> 00:00:45.300 it's a set of specific procedures for generating theory, based on qualitative data. 00:00:45.300 --> 00:00:50.610 And this view takes a couple of books as their main references. 00:00:50.610 --> 00:00:56.080 And then if you follow the advice on these books, then you're doing grounded theory. 00:00:56.080 --> 00:01:02.320 If you are building theories from qualitative data using some other ways, 00:01:02.320 --> 00:01:04.820 than the ways described in these books, 00:01:04.820 --> 00:01:06.399 then you're not doing grounded theory. 00:01:06.399 --> 00:01:11.840 Which of the two ways you use grounded theory is not particularly important, 00:01:11.840 --> 00:01:15.890 as long as it is clear to your reader what you're doing. 00:01:15.890 --> 00:01:21.759 Gioia is one of the leading experts on doing this kind of research in management. 00:01:21.759 --> 00:01:27.030 And therefore, we call following his approach as the Gioia method. 00:01:27.030 --> 00:01:32.049 So what is the Gioia method about and why do we need a method named after a person? 00:01:32.049 --> 00:01:34.090 The problem with qualitative research is 00:01:34.090 --> 00:01:39.542 that there are no real standards on how to do qualitative research, 00:01:39.569 --> 00:01:42.909 there are no right or wrong answers. 00:01:42.909 --> 00:01:44.299 To understand what that means, 00:01:44.299 --> 00:01:47.609 we can take a look at what quantitative research is about. 00:01:47.609 --> 00:01:50.130 So typically, in quantitative research, 00:01:50.130 --> 00:01:51.939 we must make a causal claim, 00:01:51.939 --> 00:01:56.079 to do so we demonstrate association, direction of influence, 00:01:56.079 --> 00:01:58.079 and we eliminate rival explanations. 00:01:58.079 --> 00:02:00.579 We have standard ways of doing that. 00:02:00.579 --> 00:02:03.029 Association, we calculate some kind of correlation, 00:02:03.029 --> 00:02:06.290 or some kind of other linear or normative association, 00:02:06.290 --> 00:02:08.970 depending on the data and phenomenon. 00:02:08.970 --> 00:02:13.980 And direction of influence, we use time delay between x and y. 00:02:13.980 --> 00:02:18.010 And then we have two main strategies for eliminating rival explanations. 00:02:18.010 --> 00:02:19.690 We use regression analysis, 00:02:19.690 --> 00:02:21.460 if that's not appropriate, 00:02:21.460 --> 00:02:25.450 we have rules that we can apply to choose an alternative technique. 00:02:25.450 --> 00:02:29.610 When we validate measurements, we use factor analysis. 00:02:29.610 --> 00:02:32.050 And we calculate coefficient alpha. 00:02:32.050 --> 00:02:35.360 And that basically covers us. 00:02:35.360 --> 00:02:38.610 When we present results, there are standards on 00:02:38.610 --> 00:02:42.702 how you build your tables to show the numbers to the readers. 00:02:42.702 --> 00:02:44.856 The problem with qualitative research has been 00:02:44.856 --> 00:02:48.300 that there are really no right and wrong answers. 00:02:48.300 --> 00:02:56.130 But recently, we have had the emergence of two different templates in management research. 00:02:56.130 --> 00:03:03.010 So basically, researchers have seen that some qualitative researchers 00:03:03.010 --> 00:03:07.730 have been able to consistently publish articles in top journals. 00:03:07.730 --> 00:03:09.446 And then we think that okay, 00:03:09.446 --> 00:03:14.320 maybe these two persons or their teams are doing something correctly. 00:03:14.320 --> 00:03:16.930 And maybe that is something that we could emulate. 00:03:16.930 --> 00:03:19.810 These approaches are, 00:03:19.810 --> 00:03:26.460 the Eisenhardt method of multiple case studies, that dates back to 1989 article, 00:03:26.460 --> 00:03:28.887 and Gioia method for grounded theory, 00:03:28.887 --> 00:03:33.380 typically a single case study that is a bit more recent. 00:03:33.380 --> 00:03:37.440 Let's take a look at the Gioia method in more detail. 00:03:37.440 --> 00:03:39.510 The differences between these methods is that 00:03:39.510 --> 00:03:43.590 the Gioia method generally applies an interpretive lens. 00:03:43.590 --> 00:03:47.040 So instead of focusing on comparing cases, 00:03:47.040 --> 00:03:49.850 how the reality of the those cases differ, 00:03:49.850 --> 00:03:54.660 and is there any patterns that allow us to make generalizable statements, 00:03:54.660 --> 00:03:56.790 that is the heart of Eisenhardts method, 00:03:56.790 --> 00:03:59.970 in the Gioia method, we typically focus on one case, 00:03:59.970 --> 00:04:04.380 and then we focus on it over time, and we interview people, 00:04:04.380 --> 00:04:05.850 we look at documents, 00:04:05.850 --> 00:04:09.730 and we try to study people's interpretations of the events, 00:04:09.730 --> 00:04:13.980 and how those interpretations shape the future of the organization 00:04:13.980 --> 00:04:16.540 or whatever you need that we are studying. 00:04:16.540 --> 00:04:20.180 So, these are different in their philosophical assumptions 00:04:20.180 --> 00:04:24.560 and they are also different in how the data are analyzed and how the data are presented. 00:04:24.560 --> 00:04:28.200 And of course, if you want to get objective facts, 00:04:28.200 --> 00:04:31.130 or if you want to get interpretations of people, 00:04:31.130 --> 00:04:35.520 you need to adjust your interview protocol accordingly. 00:04:35.520 --> 00:04:38.630 Let's take a more closer look at the Gioia method. 00:04:38.630 --> 00:04:42.450 And here is the end of the table, 00:04:42.450 --> 00:04:46.060 that shows that there are differences in how we write, 00:04:46.060 --> 00:04:48.080 when we do these tables. 00:04:48.080 --> 00:04:51.900 So the Eisenhardt report research, 00:04:51.900 --> 00:04:55.790 Eisenhardt method uses tables, tries to be objective. 00:04:55.790 --> 00:05:01.460 And then tries to basically quantify and abstract out the informant. 00:05:01.460 --> 00:05:04.450 And then the Gioia method is more about telling stories of 00:05:04.450 --> 00:05:07.850 how the informants perceive the situation. 00:05:07.850 --> 00:05:11.990 So how does one do the analysis part of the Gioia method? 00:05:11.990 --> 00:05:14.490 And how does one present the results? 00:05:14.490 --> 00:05:18.230 Well, this isn't a right or wrong way. 00:05:18.230 --> 00:05:21.710 But this is how Gioia does it, and he has been successful. 00:05:21.710 --> 00:05:24.520 Therefore, people have started emulating him. 00:05:24.520 --> 00:05:29.030 The Gioia method, the analysis part, once you have collected your data, 00:05:29.030 --> 00:05:31.609 or of course, in a qualitative analysis project, 00:05:31.609 --> 00:05:35.450 you always analyze the data while you collect it. 00:05:35.450 --> 00:05:36.790 But it starts with open coding. 00:05:36.790 --> 00:05:42.170 So the idea is that we approached the data with basically a clean slate, 00:05:42.170 --> 00:05:44.270 we don't really have an idea of, 00:05:44.270 --> 00:05:47.380 what are the key concepts that we want to derive from the data, 00:05:47.380 --> 00:05:49.070 what are theories that we want to test, 00:05:49.070 --> 00:05:53.875 rather, we just go through the interview transcripts, 00:05:53.875 --> 00:05:57.610 or whatever qualitative data we have, and then we we code it. 00:05:57.610 --> 00:06:01.560 So we try to find these lower level meanings. 00:06:01.560 --> 00:06:07.930 For example, here, these quotes, we need to listen to clients, 00:06:07.930 --> 00:06:11.770 tells us that clients are important for this company. 00:06:11.770 --> 00:06:15.360 And we need to extend beyond solving problems that clients tell us. 00:06:15.360 --> 00:06:17.360 And that basically tell us the same thing. 00:06:17.360 --> 00:06:23.400 So both of these quotes tell us that the company views clients as important. 00:06:23.400 --> 00:06:30.199 And once we have done coding of these first order categories, 00:06:30.199 --> 00:06:33.550 so we need to get the lowest level of coding, 00:06:33.550 --> 00:06:37.520 we start to combine them into more abstract categories. 00:06:37.520 --> 00:06:42.250 And this produces something that Gioia calls the data structure. 00:06:42.250 --> 00:06:44.850 And we have the title of your data structure. 00:06:44.850 --> 00:06:48.430 So the first order categories, these are the codes, 00:06:48.430 --> 00:06:51.720 you might have a couple of 100 codes in the big project. 00:06:51.720 --> 00:06:55.510 So this is basically a representative sample of codes. 00:06:55.510 --> 00:06:57.820 And then you start to combine codes, 00:06:57.820 --> 00:07:01.949 for example, need to be more market pool clients are more important, 00:07:01.949 --> 00:07:04.780 that creates an identity change imperative. 00:07:04.780 --> 00:07:10.170 So this company is more technology focused company, and now they need to become 00:07:10.170 --> 00:07:13.610 more customer, more market focused company. 00:07:13.610 --> 00:07:19.750 So we are combining these first order categories into higher order themes. 00:07:19.750 --> 00:07:22.120 And then once we have the higher order themes, 00:07:22.120 --> 00:07:26.734 we are still increasing the level of abstraction 00:07:26.734 --> 00:07:30.949 by combining them under overarching dimentions. 00:07:30.949 --> 00:07:36.919 For example, we have new knowledge imperative that have three different dimensions. 00:07:36.919 --> 00:07:39.240 Once we have completed the data structure, 00:07:39.240 --> 00:07:41.200 which means that we have the coding, 00:07:41.200 --> 00:07:43.320 and we have combined the categories, 00:07:43.320 --> 00:07:47.730 there are the initial first level categories into second order categories, 00:07:47.730 --> 00:07:49.370 and then overarching dimensions, 00:07:49.370 --> 00:07:52.220 then, we start to build our theory. 00:07:52.220 --> 00:07:54.195 And typically in Gioia's approach, 00:07:54.195 --> 00:07:59.300 the theory is presented as a process that unfolds over time. 00:07:59.300 --> 00:08:02.170 And he likes to use this kind of graphics. 00:08:02.170 --> 00:08:04.169 So there is pre-graft context, 00:08:04.169 --> 00:08:08.880 he tells what was the company before the knowledge graft attempt, 00:08:08.880 --> 00:08:13.490 then what happened when the knowledge graft was applied? 00:08:13.490 --> 00:08:14.880 What was the implications? 00:08:14.880 --> 00:08:17.140 And how did people respond to that? 00:08:17.140 --> 00:08:19.230 And then you construct a theory for 00:08:19.230 --> 00:08:24.360 why this happened based on people's interpretations. 00:08:24.360 --> 00:08:30.570 There are some guidelines on how to do this kind of research well. 00:08:30.570 --> 00:08:34.392 And these are guidelines presented by Gioia himself 00:08:34.392 --> 00:08:37.290 in an Organizational Research Methods article. 00:08:37.290 --> 00:08:42.760 And he says that this research starts with a good research question. 00:08:42.760 --> 00:08:46.440 So the research question is typically a more open ended than 00:08:46.440 --> 00:08:49.580 would be in Eisenheardt's approach. 00:08:49.580 --> 00:08:53.930 And it might not rely as much on prior theory, 00:08:53.930 --> 00:08:57.740 as much as it relies on an interesting pheomenon. 00:08:57.740 --> 00:09:00.820 And then you want to understand the phenomenon. 00:09:00.820 --> 00:09:03.440 And you basically start from a clean slate, 00:09:03.440 --> 00:09:06.600 then you are you collect rich data from the informants, 00:09:06.600 --> 00:09:10.860 and particularly focus on how the informants interpret the situation. 00:09:10.860 --> 00:09:15.570 Then you analyze the data based on the three levels of abstraction. 00:09:15.570 --> 00:09:20.690 And then you collect a narrative that explains what happened in the organization and why. 00:09:20.690 --> 00:09:24.990 And then finally, you write your report.