Due to an increasing interest in the use of single-subject experimental designs (SSEDs), appropriate techniques are needed to analyze this type of data. The purpose of this paper proposal is to present four studies (Beretvas, Hembry, Van den Noortgate, & Ferron, 2013; Bunuan, Hembry & Beretvas, 2013; Moeyaert, Ugille, Ferron, Beretvas, & Van den Noortgate, 2013c; Petit-Bois, Baek & Ferron, 2013) each studying empirically one particular extension of the basic two-level or basic three-level model that can be used to synthesize SSED data across cases (i.e., two-level model) and across cases and studies (three-level model). Each extension is discussed and examined through intensive simulation studies by looking at the bias and precision of the parameter estimates, and the validity of statistical inferences. The results indicate whether the extended two- and three-level models are recommended to summarize SSEDs and in which conditions. This is a timely issue, since the number of published SSEDs is increasing rapidly together with the need for adequate techniques to analyze and meta-analyze SSEDs answering the same underlying research question. The authors conclude that the multilevel model is flexible enough to be extended and is found to be appropriate for combining SSEDs. However, there are still a lot of other extensions that are not discussed here that are being investigated and that still remain to be investigated such as other estimation procedures (i.e., bootstrapping and Bayesian estimation), count outcomes, and multiple dependent outcome variables.