Eval C3

Resource link

EvalC3 is an organised set of tools for developing, exploring and evaluating predictive models of expected outcomes.

This resource and the following information was contributed by  Rick Davies. 

Authors and their affiliation

Rick Davies, Cambridge, UK

Key features

EvalC3 is an Excel app, that contains a set of macro-based tools that enable the user to work through a series of relatively simple data analyses, organised into a simple workflow, with a data set that is of interest to them. It is available for free and is supported by a website providing detailed information on each stage of the workflow.

It brings together ideas from two different fields:

(a) simple search algorithms from the field of machine learning and

(b) a usable view of complex causality from Qualitative Comparative Analysis (QCA).

Most of the workflow is focused on cross-case analysis, but the analysis outputs lead the user to case selection strategies which can then be the basis of within-case investigations. The latter is needed in order to identify if good performing predictive models can also be useful explanatory models.

The EvalC3 analysis process begins with the pasting in of a copy of a data set that has cases as rows and columns as attributes of those cases, including outcomes observed for those cases. Cases can be households, individuals, communities, organisations, governments, or nations. But in any data set they must be all of one kind i..e not a mix of these types. Each cell in the data set has binary (1/0) value telling us if the row case has the column attribute/outcome, or not.

The result of EvalC3 analyses is usually a set of predictive models, describing under what conditions an outcome of interest is present (or absent). For example that where x and y are present then outcome z will be present. Or, the combination of A and B is sufficient for the outcome to be absent. or, the combination of Y and P is necessary for the outcome to be present

Free skype based technical support is available for all users

How have you used or intend on using this resource?

A range of types of possible uses includes during project selection, during project implementation, during a project evaluation and during a review of existing evaluation. 

Sources

Davies, Rick. (2018). EvalC3 – …tools for developing, exploring and evaluating predictive models of expected outcomes. Retrieved 28 August 2020, from https://evalc3.net/