... In our case, we tested our data with different threshold levels and finally arrived at two citations per year, with a stress value of 0.0742 indicating a good fit. 2 The algorithm proceeds in the following steps:Consider every data point (articles in our case) as individual cluster.Calculate the similarity of one cluster with another cluster (calculate proximity matrix) based on standardized co-citation value.Merge clusters which are highly similar or close to each other.Recalculate proximity matrix for each cluster.Repeat steps 3 and 4, until only a single cluster remains. 3 List of core documents: Battilana and Dorado (2010), Battilana and Lee (2014), Besharov and Smith (2014), Binder (2007), D 'Aunno et al. (1991), DiMaggio and Powell (1983), Dunn andJones, 2010, Friedland andAlford (1991), Glynn and Lounsbury (2005) (2008), Lawrence and Suddaby (2006), Lok (2010), Lounsbury and Crumley (2007), McPherson and Sauder (2013), Meyer and Rowen (1977), Oliver, 1991, Pache and Santos (2010, 2013a, Purdy and Gray (2009), Rao et al. (2003), Hinings (2005, 2009), Scott (2001), Scott et al. (2000), Seo and Creed (2002), Smets et al. (2012), Suddaby and Greenwood (2005), Thornton (2002Thornton ( , 2004, Thornton and Ocasio (2008), Thornton et al. (2012), Tracey et al. (2011), Zilber (2002. ...