Evaluation Models

Analytical models

It is related to epistemological and explanatory paradigms.

Provides explanation based on related theories

Studies observable phenomena

Scientific credibility is based on internal validity.

Uses quantitative information

Follows the deductive method

Privileges statistical analysis

Uses measurement instruments such as the survey

Establishes rules of behavior

In the models are

Cost-benefit

It is based on the quantification and comparison of updated costs and benefits.

Comparison, indicators such as VAN, TIR, among others, emerge.

Cost effectiveness

Evaluates whether the project generates real effects on the target population, beyond the economic benefit.

Selects the most economically efficient alternative, comparing unit or total costs for the same benefit.

Global models

Are based on interpretive paradigms

Among them we can distinguish

CIPP Model

Provide useful information at all stages of the project.

Evaluated from planning to final results

Focused utilization model

Involve leaders and users of information throughout the process.

Choose methods according to utility

Illuminative evaluation model

Describes how the actors live and interpret the project

Fits the meanings that emerge during the evaluation

Expert evaluation model

It is based on the evaluator's accumulated experience and critical sensitivity.

Uses concepts such as referential adequacy and structural corroboration

Formalized models

They are based on the systematization and rigorous control of variables and seek internal validity and replicability.

Among them we can distinguish

Experimental Model

The evaluator manages the experimental conditions

Uses an experimental group and a control group in closed settings e.g. hospitals, schools, etc.

Quasi-experimental Model

No random selection of groups

Comparable but not equivalent groups are worked with

Regression Model

Use statistical techniques to control variables

Requires a robust database and a sound theoretical model

Semi-formalized models

They seek to observe the historical behavior of variables and compare them with the project and analyze changes over time

These include

Non-experimental before and after model

A variable is measured before and after the intervention

If there is no previous measurement, it can be estimated with retrospective questions.

Model after with comparison group

No baseline, but comparison with another group or parallel intervention.

Independent variables are statistically controlled at the end of the project.

Model after

Only what happened after the project is recorded, without prior measurement

Aplicable a proyectos comunitarios, ONG o gobiernos locales donde no hay línea de base.

Instruments and techniques in the evaluation models