Measurement Types: Guess, Estimate, or Direct Measurement?
This article explains the three measurement types in the report builder and how to choose the right one for each indicator.
For each outcome indicator in your impact model, you will be asked to describe how you measured it. There are three primary buckets with relevant sub-categories:
Direct Measurement You directly tracked and measured this outcome. This includes formal assessments, pre/post surveys, administrative data, and self-reported data collected from participants (such as surveys asking participants whether their situation improved).
Example: You surveyed program participants before and after completing your program and tracked the percentage who reported improved outcomes.
Evidence-Based Estimate You estimated this outcome based on data from other similar programs, government or sector-wide research, or your organization's past results.
Example: You used national statistics on program effectiveness to estimate the proportion of participants likely to achieve a specific outcome.
Guess You do not currently track or measure this outcome directly and are making an educated assumption based on experience or logical reasoning.
Example: You believe your program improves participant well-being based on anecdotal feedback, but you do not have data to support a specific number.
Which should you choose?
Use whichever option most accurately describes how you arrived at your numbers. Choosing Direct Measurement or Evidence-Based Estimates where possible will strengthen your report and reflect well on your organization's measurement practices.
If you find yourself selecting Guess often, that is useful information and it can help you identify where to invest in better data collection over time. See Improving Data Quality and True Impact's Top 5 Data Collection Best Practices for practical guidance on strengthening your measurement approach.
A note on using percentages
The report builder asks for the number of individuals who experienced each outcome, not percentages. If you have percentage-based data, multiply the percentage by your total participant count to get a number. For example, if 80% of your 200 participants improved an outcome, enter 160.