This project is now complete. The published manuscript appears in *Journal of Social Psycholog* and can be accessed [here][1]: * Calin-Jageman, R. J. (2018). Direct replications of Ottati et al. (2015): The earned dogmatism effect occurs only with some manipulations of expertise. Journal of Experimental Social Psychology, 1–10. doi: [10.1016/j.jesp.2017.12.008][2] Ottati et al. (2015) reported a series of studies supporting the *Earned Dogmatism Effect*, the notion that feeling like an expert makes one feel more entitled to be close-minded. The goal of this project is to complete high-powered, precise, and pre-registered replications of ~~two~~ ~~three~~ four key studies from Ottati et al. (2015): * Study 3 - In which perceptions of expertise were manipulated via participants completing an easy or difficult retrieval task (remember 2 or 10 policies enacted by President Obama). Those who completed the easy task scored higher on a measure of political open-minded cognition. * Study 4 - In which perceptions of expertise were manipulated via participants completing an easy or difficult trivia test. Those who completed the easy test scored higher on a measure of open-minded cognition. * Study 6 - In which participants envision themselves in a social situation with others who know much less than them or about the same (counterbalanced order). Open-minded cognition was measured after each scenario, and was found to be lower when envisioning interacting with those who know much less (there's a good bit more to this complex study, which also measured control variables and normative-entitlement for another person interacting with novices or equals). * The "alternative mediators" study in which Ottati et al. examined if manipulation of expertise alters self-esteem, sense of power, or mood, to determine possible counfounds/mediators of the earned dogmatism effect. In the original study, only the success/failure via trivia test was examined, with participants completing the procedure of Study 4 but with the alternative mediators measured rather than OMC. The goal of the additional replication is to repeat this study with higher sample size and also to test the imagined scenarios manipulation (Study 6) in the same way to determine if one or the other type of manipulation elicits different changes in these state variables. Materials for these studies are precisely matched to the original thanks to the gracious cooperation of the original research team. ## Current Status Complete. Manuscript written and submitted for review, all data posted. * Study 3 - Completed replication attempts with a psych participant pool and with an online US sample (MTurk). Effect sizes are near zero for open-minded cognition. * Study 4 - Completed replication attempts with a psych participant pool, an online UK sample (Prolific Academic), and an online US sample (MTurk) . Effect sizes are near zero for open-minded cognition. * Study 6 - Completed replications with on online Canadian sample (Prolific Academic), and online UK sample (Prolific Academic), and psych participant pool. Expected effects on normative entitlement (showing expected norms about experts being allowed to be closed minded) and on OMC scores (showing participants agreeing they would be closed minded when dealing with those who know very little about a topic). * Study 6 Better Manipulation - Complete. In Study 6, participants in the expert condition are told that those they are interacting with "know very little about politics". Thus, expertise is created not only by having more knowledge but also by colleagues having almost no knowledge. This may activate not only norms related to expertise but also norms related for dealing with complete ignorance. Thus, the large and replicable effects detected with this study may not be directly related to self-expertise but instead could reflect the influence of dealing with other-ignorance. To test this idea, on online MTurk sample was collected. Participants were randomly assigned to the same type of expertise manipulation (others know very little) or to a relative manipulation (a topic you happen to know a whole lot about). It was found that this variation in how expertise was manipulated did not interact with the expertise effect--either type produced a similar decrease in open-mindedness in the expertise scenarios. * Study 4 Alternative Mediators. Complete. One possible explanation for strong effects in the within-suubjects design yet no effects in the success/failure manipulations is the possibility that the success/failure manipulation alters more than just expertise--it may alter mood or thinkign in ways that can counteract the earned dogmatism effect. Ottati et al tested this by having participants complete the same easy/difficult trivia test as in study 4 but then measuring various affective states, state self-esteem, and state power. They found siginificant effects for some affective states, though this was found with a small sample size and no corrections for multiple comparisons. Thus, these finding may not be robust, and this explanation of the differences might be specious. To help sort this out, the last step of this project is to repeat this examination for additional changes in affective states, etc. * Study 6 Alternative MEdiators - about to begin. A key part of the explanation for the differential effectiveness of the imagined scenarios manipulation (Study 6) compared to the success/failure manipulation (Studies 3 and 4) is that they may evoked different changes in state variables. To fully evaluate this, then, it is necessary to measure the same state variables with the imagined scenarios manipulation as well. ## Registration History Replication attempts are made [pre-registered][3] using the OSF pre-registration template. The pre-registration plan specifies sample sizes, research questions, analysis strategies, etc. In general, the plan specifies mirroring the original research as closely as possible. An initial overall plan was developed for both Study 4 and 6. Then, as each replication has completed, the next stage of the plan was refined/expanded and re-registered. - [1st pre-registration][4], 2016/01/12 -- completed for pre-registration challenge, describes plans for in-person replication of Study 2 and online-replication of Study 6. - [2nd pre-registration][5], 2016/01/25 -- completed to lock in plan to conduct replications sequentially, completed just before in-person replication of Study 2. - [3rd][6] and [4th pre-registration][7], 2016/02/29 - completed prior to start of online replication for Study 6. Plan to recruit Canadian participants via Mturk did not work out (not enough of these users left, 0 participants accepted the HIT in first few hours). Switched to Prolific Academic and a payment of 1 Euro (required by PA to meet their hourly minimum). - [5th pre-registration][8] 2016/03/03- completed prior to additional replications of Study 6 with a psych participant pool (50 participants planned) and a UK sample from Prolific Academic (50 participants planned). Also added an additional replication of Study 4 with a Prolific Academic sample (125 participants planned). - [6th pre-registration][9] 2016/03/10- completed prior to additional replication of Study 4 with MTurk pool. Locks in familiarity screen (1-4, those scoring 3 or 4 excluded) and sharpened suspicion check, which will help exclude those who have already participated in the original research and/or have heard about the original research in the media. The UK prolific academic studies planned in the 6th pre-registration are completed; the additional replication of Study 6 with a psych participant pool is ongoing. - [7th pre-registration][10] ~~2016/09/10~~ / 2016/11/17- added attempt to replicate Study 3 with a psych participant pool. Meant to register prior to data collection, but failed to click the final submit button! - [8th pre-registration][11] 2016/11/17 - added attempt to replicate Study 3 with an MTurk sample. - [9th pre-registration][12] 2016/11/28 - added to attempt to replicate Study 6 with a better manipulation. - [10th pre-registration][13] 2016/11/30 - added to attempt to replicate the "Study 4 Alternative Mediators" study. - [11th pre-registration][14] 2017/03/21 - added to complete the Study 6 Alternative Mediators study. - ## Differences from Original Studies In what ways were these replications different? * Study 3 replication is being conducted with a psych participant pool rather and with online US sample (MTurk). A positive control and suspicion check was added. For MTurk a familiarity check was added. * Study 4 was conducted with a psych participant pool, online UK sample (prolific academic), and with a US online sample (MTurk); the original was done with an online US sample (MTurk). A positive control and suspicion check was added. * Study 4 used trivia questions from middle-school and collegiate academic bowls. Some questions, though, were very specific to the U.S.A. For the UK sample, those that were specific were replaced by general questions from the same academic level. * Study 6 - Was completed with a Canadian sample of Prolific Academic participants, a UK sample of Prolific Academic participants, and a psychology participant pool. The original study used U.S. workers on MTurk. However, due to IRB constraints we could not obtain a list of prior MTurk participants, so using the same country could lead to duplicate participants. * Study 6 Adapted. By design, two different ways of manipulating expertise were used. In addition, the scenarios about John to measure normative entitlement were dropped as were the measure of state power and of affective state. These were dropped because these research questions were not of interest for this replication. In addition, the scenarios were altered slightly to be about a specific political issue rather than politics in general to make it more plausible that a participant could imagine knowing a great deal or an average amount within the conversation taking place. * Study 4 Alternative Mediators - In the original study attitude certainty and extremity were measured. These have been dropped because no significant differences were found and because the scales don't seem to really apply to the study. With the original researcher's permission, materials for these replications have been posted. [1]: http://linkinghub.elsevier.com/retrieve/pii/S0022103117302913 [2]: http://linkinghub.elsevier.com/retrieve/pii/S0022103117302913 [3]: https://osf.io/5f9eg/ [4]: https://osf.io/5f9eg/ [5]: https://osf.io/87n6d/ [6]: https://osf.io/4zm5a/ [7]: https://osf.io/ftzw2/ [8]: https://osf.io/3qg2a/ [9]: https://osf.io/9268j/ [10]: https://osf.io/z7gsr/ [11]: https://osf.io/z8vd9/ [12]: https://osf.io/tv4tz/ [13]: https://osf.io/t4equ/ [14]: https://osf.io/tbd