Target Groups and Sampling
Select each chain link to learn more
Defining
the Question


Defining the Question
The very first link in the data chain is to correctly define the question we want our data to answer—that is, what are we trying to measure?


Designing a
Measurement Strategy
Designing a Measurement Strategy
The second link in the data chain is to create a data collection and measurement strategy that fits our question.
Target Groups
and Sampling


Target Groups and Sampling
The third link in the chain is to identify the population—that is, the target group—we want to measure and make sure that the population we are measuring is actually the population we want to measure. Sampling is always a crucial issue, whether we are taking a subset of our target group or attempting a census—that is, trying to contact everyone in our target group. If our sample differs from the larger group of people we want to say something about, then the answers we get from our data will not be true about the larger group.


Collection
and Entry
Collection and Entry
Once we establish the sample, we must collect and enter the data. Accurate data collection and entry are extremely important. Poor data entry or missing data can generate flawed results.
Preparation
and Analysis


Preparation and Analysis
Staff must prepare data they gather from multiple sources for analysis. This preparation includes making sure categories, names, and formats are standard across data sets from different sources. Preparation also involves dealing with missing and duplicate data and merging or extracting groups of records to preserve confidentiality.
Once staff have prepared the data, there are many decisions they need to make related to analyzing the data. For example, they need to decide.
- what kind of analysis will give the desired results;
- how they should define data elements;
- when and how they should calculate new variables;
- what thresholds they should use to put individuals into categories; and
- should they express their results as totals or percentages?


Interpretation
Interpretation
The other links in the data chain predetermine interpretation of the data. For example, if we have designed our data chain so that it can answer a question about comparisons between groups, then we can interpret what it says about comparisons between groups—otherwise, we can’t.
Dissemination
and Use


Dissemination and Use
Data dissemination includes both distribution and data visualization; the data must be sent out into the world and in a format that honestly represents the meaning of the result and is accessible to the end users.
- ‹ previous
- 16 of 41
- next ›