 Ja jos se on hyvää rikkoa, olen pystynyt tämän YouTube-yhtiön kanavalle. Sitä voidaan tehdä jotain, jotain, jotain, jotain? Joo, joo, joo, joo, joo. Tämä on tärkeintä, miten se vaikuttaa. Jos haluan muistaa paljon, niin haluan vain... Haluan vain puhua online ja... ...unlistaa videon vain, johon olen pystynyt. Noin, nyt on aikaa. Ja... Olen huomattu, että olen tullut 7 vuotta, mutta olen tullut niin, että katsotaan. Ja sitten olen tullut katsomaan Reza ja Reza, josta olen tullut ensimmäinen kautta, mutta hän oli visa-asio, joten hän ei tullut tullut tullut tullut. Tällä kertaa, että katsotaan, että katsotaan, että katsotaan, että katsotaan, että katsotaan. Ja hän katsotaan, että hän tullut tullut katsotaan, että katsotaan, että katsotaan, että katsotaan, että katsotaan. Ja sitten hän tullut se, että katsotaan Reza ja hän tullut ensimmäinen kautta katsotaan Reza-asioon, joka on vain ei ole helppoa. Täällä olemme, mutta olen tullut se. Ja tämän kertaa on entisempilä-aurintaa, ja olemme taloutuneet seuraavalla ja measurements-asioon. Parti on aina, mitä Reza on ja täysin mennä. Mietin se, että tämän lopullekijöiden hänä on tosiaan residence-asioon ja hän paitaa Reza-asioon. Kyllä me olemme taloutuneet seuraavalla ja measurements-asioon, koska hän ei tullut paremmin tiedä asioon laittamista, vuonna olet tapauksessa, että hän hän hänesää. Tämä on vähän problemaattista, koska olemme nähneet, että on missä matkaisuja, mitä on EO-konsektua ja miten se on oikeastaan ympäristö. Ja sitten on vähän asioita, että ympäristö, prosessuus ja modelen prosessuus on se, että se on tärkeintä asioita. Joten, mitä on EO? EO on tietysti ympäristö, johon se on tärkeintä asioita, johon se on tärkeintä asioita, johon se on tärkeintä prosessuus, innovatioita ja riske-takka, tai se on ollut ympäristö, johon se on tärkeintä asioita, mutta on missä tärkeintä asioita, prosessuus, innovatioita ja riske-takka. On myös 5-tärkeintä asioita, mutta me olemme vain nämä 3-tärkeintä asioita, kaikki, että sanoin, on myös ympäristö, mutta, tietysti näin, johon se on tärkeintä asioita, johon se on tärkeintä asioita, esim. Koska jät kardeşim viimeista seikkailua, jaнаam Victorian ryhmistä, formarilya tai mon drovech wordt eluvelliselt. weeks, but we have had independent effects on outcomes, so innovative effect of performance, independently of the evolution. And we measure the model, and as a component of the EEO, we are confounding the EEO concept with the independent effects of the models. Let's take a look at the background of the paper. I'm just going to skim through the background. The idea of EEO was originally that it was kind of like a label for a collection of attributes. So it started from Miller's 1984 paper where he was just looking for indicators or correlates of entrepreneurship. So he was not thinking about a concept that exists on its own, but simply a collection of trades. Like you might think about a talented person who would have certain trades and talent itself might not be something that you are interested in, but you're just referring to this collection. And then after, you know, in somewhere late 80s, early 90s EEO started to have kind of like theoretical existence of its own independently of these dimensions. So that's the background, and there are three conceptual issues about how we define EEO that we talk about in the paper. The first problem is that how we understand entrepreneurship has evolved over time, but EEO has not kept up. So if we call something entrepreneur orientation, then it must be related to entrepreneurship, otherwise we are just confusing everybody. But entrepreneurship, what it refers to, has been shifting. So we resurrected these various things that entrepreneurship has been identified with. And when Miller did his work, entrepreneurship was about creating new organizations, entering new markets. It was more about these actions and maybe some kind of processes. But nowadays we look at it instead of a process or a trade, we look at entrepreneurship more as a practice. So the conceptualization of EEO of entrepreneurship has shifted. We look more at opportunities instead of new organizations or new entries to new markets. And EEO has been basically the same three dimensions since Miller's work and his does not kept up. So entrepreneurship has been going there and EEO is going there. So we have an increasing disconnection between these two concepts. Then we talk about our dispositional versus behavioral aspect of EEO. And this was highlighted in Anderson's 2015 paper where he talks about EEO as containing two dimensions of behaviors and dispositions which contain attitude towards risk and behaviors contained proactiveness in the way it is. The problem with this dispositional versus behavioral is that dispositions and behaviors are not the same. So they are causally related. So we have dispositional EEO and behavioral EEO that are both referred to as EEO. And this causes confusion because then a reader of a paper about EEO would not know whether that paper is about behaviors or whether it's about dispositions. So there is referring to both cause and effect using the same term is potentially confusing. Then there is the third question which is perhaps the most important conceptual question in our paper. And it is the link between EEO and innovativeness. And this relates to the area of philosophical meteorology which relates to part-hole relationships. And there are two extremes that we can think of. One is that the EEO is simply a convenient label for innovativeness, risk taking and proactiveness and nothing more. EEO is simply a sum of these parts and then the other extreme is EEO exists independently of these dimensions and these dimensions only exist as measures. EEO started basically from here and it has been shifting more to the right but still the measures we apply for EEO are the same. If we apply innovativeness, risk taking and proactiveness as measures of EEO, this causes a problem because not all innovativeness is entrepreneurial, not all entrepreneurship is innovative. So innovativeness has effects that are unique, not depending on entrepreneurship. Entrepreneurial orientation has effects that are not depending on innovativeness. So if we have innovativeness as measure of entrepreneurial orientation, we are confounding two different things. The same argument applies to risk taking and proactiveness. So there are three conceptual issues. One defining EEO through these traits has not kept up with how evolving conceptualization of what entrepreneurship is. So we think about entrepreneurship differently now but our thinking about EEO has not kept pace. Defining EEO as this positional concept we have here concept is confusing and it's not according to good scientific practice where one label should refer to one concept. It is confusing to refer use the same label or same concept for two different meanings depending on the context. Innovativeness, proactiveness, risk taking are problematic as measures because they have independent existence of EEO. Of course not all researchers define EEO in terms of these dimensions but they are still used as measures. Now we go to the main part of the paper which is empirical issues in EEO measurement slash modeling. There is currently a debate and has been a debate for some years about whether EEO should be modeled as reflectively or formatively. The idea of this reflective model is that a firm is innovative, proactive and risk taking because it is entrepreneurial. Being entrepreneurial courses are your company to be high on innovative and proactive risk taking and then in the formative or conceptualization or formative modeling the idea is the firm is entrepreneurial because it's innovative, proactive and takes risks so the direction of model causality goes in the different ways. The reflective model has received lots of criticism lately so this is not realistic because it assumes that the dimensions are highly correlated. They are not empirical correlated. They are expected to share the same causes. That's not very realistic. Innovativeness has different causes than risk taking for example. They are not uniformly caused by the same things. The formative model is gaining popularity but there are issues that we talk about in this paper that have gone unnoticed in the formative modeling of EEO. We need to understand why this became popular. The popularity of formative approach can be attributed to two main papers. There is George 2011 in a Journal of Management if I remember correctly. He did a simulation study where he showed that if you mis-specify a reflective model where a formative model is correct your estimates are biased by two and a half fold which is pretty big bias. Then Anderson continued on that and refers to these as nomological errors. They do an intricate study where they show that the data do not support the use of reflective model and they strongly recommend the formative model. But there is a bit of a but to George's argument. It was shown by Agura Uretan Maracas in 2012 using a different context that this bias is actually just a methodological problem related to how latent variables are scaling. I will not walk you through the argument. We present it in the paper and we replicate George's paper and we show that this bias dimension just goes away basically when you consider standardised estimates which have the same scale between two samples. So this is something one problem we talk about in the paper. Then there is another problem called interpretation confounding. This is discussed in the methodological literature but it's not discussed in the EOL literature. The problem is that when this kind of formative model we estimate these weights so innovator is productive and risk taking we estimate them for each sample for each study separately. So we would think that our original EOL was a sum of these items but nowadays we actually use a weighted sum. So one study might have innovator's productive risk taking all weighted equally. Another study might have risk taking weighted 10 times more than innovator's proactiveness. Another study might have innovation weighted 10 times more than proactiveness and risk taking. So are these studies really focusing or talking about the same thing. In one study EOL is innovativeness 90 percent and other study EOL is 80 percent risk taking. Do these talk about the same things empirically. Then of course we might have scenarios and we do have scenarios where risk taking particularly has a negative effect. So one study says that EOL is positively related to risk. They define EOL as positive correlate of risk. Another study defines EOL as a negative correlate of risk. So what do we expect when we have this kind of models. Another thing that has been not addressed in the EOL literature that is discussed in the methodological literature comes from Rigdon's research. That's not cited on the slides but he explains this really well. The paper is that if we have the formative conceptualization innovativeness proactiveness risk taking. They cause EOL which then causes different outcomes. This model assumes that the effects of innovativeness proactiveness risk taking are always at the same proportions. So if risk taking is twice as important determinant of some outcome compared to innovativeness. Then it must be twice as important determinant of any other outcome. That does not sound very realistic. And this is something that means also that the dimensions are interchangeable this model. So it means that if you lack innovativeness you can compensate for by increasing your risk taking. And that's because EOL is measured model as weighted sum. You know from high school from elementary school math that if you take a sum of two things. If you lack one thing you can increase the other thing and that increases that makes it some of the same. So we build on Anderson's argument who says that EOL dimensions don't co-vary in a way that would support the reflective model. We argue that they don't co-vary with the outcomes in a way that would support the use of the formative model. And we test this with meta-analysis. So what we do in the paper is that we collect various studies and we collect the effect sizes of these three different dimensions. Of those studies that study the effects of innovativeness risk taking and proactiveness separately. We run a model, we test the hypothesis that the proportions are constant across studies. This hypothesis is strongly rejected. So the effects are not the same for every possible outcome which is pretty obvious. But we need to support it with data. Then there is a third problem. So we had a problem of interpretation of confounding. We had a problem of full media so it's not supported. And we have this interesting problem still called identification problem. I will not talk about the technical issue here. But this model requires that you have two reflective indicators for the formatively measured latent variable. The problem is, and this comes from Aguira Oretas and my research published in measurement. The problem is that in this case the empirical meaning of entrepreneurial orientation does not come from these dimensions at all. It comes from these two indicators. We show this by replicating Andersson's paper. So we generate data set that has the same covariance that they have. So it produces the same results. So we reproduce the results using synthetic data. And then we test what happens if we have two reflective indicators and two these formative dimensions that are not related. So does the result of EO on performance, does that effect depend on how the EO dimension is related performance? Or does it depend on how these identification indicators determine performance? We show that if these two indicators that we just add for identification purposes, if they are not related to performance at all, all performance effects disappear. So it means that these behaviors or these effects of the dimensions and EO on performance, they disappear because EO is not in this model defined as some of these dimensions. Rather it's a latent variable with two reflective measures. We also show by replicating another scenario where EO dimensions are unrelated to performance. And we show that when performance is considered, if the indicators that are used for identification on related performance, there is no effect. If EO dimensions are not related performance, there is no difference. We can also freely leave out all the EO dimensions. And the result will not change. So we leave out no EO dimensions. We leave under bear behaviors away. We leave attitude away from the model. The result does not change. It shows that these kind of models don't measure the latent variable through the dimensions. Okay, so we showed a couple of problems in the current practice in modeling EO in the paper. So what do we do with under preorientation? We have a couple of key arguments. And this is something that we are still working on before we submit this to the journal. Our key claim, our key recommendation is that we should stop forcing these dimensions into a single latent variable because they don't have the same causes and the same consequences. Innovative product is a risk taking, our separate variables and they should be modeled separately. The next thing that we think about is maybe we should retire the concept to EO. So if we have EO and if we use EO to refer to organizational level entrepreneurial behaviors, why don't just call them organizational entrepreneurial behaviors, our organizational intelligence systems instead of EO, which is pretty vague as a term. And then we think also that maybe we should go back to the rules to Miller's work and use EO simply as a convenient index. So sometimes you have so many variables and maybe entrepreneurship is a minor thing in your paper that an index of these dimensions might be a crude measure like Miller did. But calling this as EO would be a bit problematic because the term has some theoretical meaning. So maybe we should just then call it as entrepreneurship index instead of EO. Thank you. Any thoughts or questions? Do you have a hunch how this could be? Well, you talked about now a mixture of empirical and conceptual things. When we now take a look on what has been done so far, so is there anything that we could do better? For instance, when we now, when we replicate the study that has been done already, or only empirically speaking, so would it be helpful to model a second latent factor, an unmeasured latent factor, something like that? This would it be helpful to, I don't know, to loosen some covariance between the indicators of EO? Would it be, do you have any ideas in this regard or is that simply meaningless? I'm not saying that EO's meaning, as EO clearly has meaning as a theoretical concept, but the measurement approach is problematic. And if you have the three measures, innovativeness, productivity, and risk taking, I would think that it's very important to understand under which conditions each of those has the positive or negative effects instead of lumping them together as one thing. This builds on Edward's argument in 2001 in the organization research methods, where he points out that typically when field matures, we start with these very abstract concepts and then we start modeling dimensions. And EO is pretty mature already, so maybe we should start moving forward, looking at individual dimensions instead of lumping them together. So that's, as replications go, I think that's what we should do. And then entrepreneurial orientation clearly needs a bit more of a conceptual refinement, like unification. There are so many different ways of defining EO and maybe we should also start talking about, like not having this dispositional EO attitude, behavioral EO, individual level EO, what not EO and just use different terms to avoid confusing. Yeah. Have you seen any papers that do this, that move away already or? Minority of papers do model in the measures separately, and that allowed us to do the meta analysis to check how much the effects vary between studies. But most studies lump them together and don't necessarily report the formative weights. So there might be a structural equation model where risk taking has a negative effect on EO, and it's called an EO that is defined through risk taking, and yet the effect is negative, which is inconsistent with the theory. We have developed the paper quite a lot since the R&M submission. If someone wants to read the most recent version, I can send it by mail. Okay. All right, yeah. Time is also up. Thank you so much. Yeah, you're welcome.