Intercultural Practice and Equity · Internal leadership

Responsible Data and Evaluation

A number is a record of what we counted, not a record of what happened. Four lessons on who reaches the count, what a category hides, how a finding can describe a community instead of a system, and the short check to run before a number becomes a decision.

40–45 minutes · 4 lessons · Voluntary learning

Start learning

Course outline

Who reaches the count
Division numbers describe the people who found the door, got through it and were recorded correctly. Everything that happens before the record is a filter, and the filters do not operate at random.
11 minutes
What the categories were built for
Every field in a record was defined by someone, for a reason, and the reason was usually operational. Trouble starts when a category built for payment or eligibility is asked a question about people's lives.
11 minutes
From a finding to a story
The same gap can be written as a fact about a community or a fact about the process that reached them. Both can be true. Only one of them names something the division can change.
10 minutes
Your data-interpretation checklist
Six questions to run before a number becomes a decision, short enough that you will actually use them, and written down so they do not depend on who happens to be in the room.
10 minutes

Share this course

What you’ll be able to do

Explore and practice

Identify the denominator behind a division number and name at least three groups the count cannot include.
Trace one category in a report back to the form or rule that defined it, and state what it records, what it groups together and what it cannot show.
Read a missing value as information about the process rather than as a default, and say what your own reports currently do with blanks.
Rewrite a finding so the process is the subject of the sentence, and name the comparison and the causal claim the original version assumed.
Draft a six-line data-interpretation checklist for one report you produce or rely on, including what happens when a line cannot be answered.

What you can take into your work

Four worked scenarios drawn from a timeliness figure with a hidden denominator, a breakdown requested from fields that were never built for it, a briefing sentence that explains a gap by describing a community, and an interpretation check that grew too large to use.
A knowledge check in every lesson with feedback that explains the usable answer.
Sorting practice that places losses before, inside and after the record, and separates a sentence about the process from a sentence about the people and a claim the data cannot carry.
A completed six-line checklist for one real report, with at least one line recorded honestly as unknown.

Take one report you produce or rely on. Write in one sentence what its headline number counts, list the groups that cannot appear in it, and run the six lines once with a colleague who did not build the number.

Read the introduction

Division numbers are made of records, and a record exists because something already went right between a person and the department. Someone told them the service existed, in a language they could use. They reached a door that was open when they could get to it. Somebody entered them correctly. Everyone who did not get that far is outside the count and invisible inside it. This module is about reading division numbers with that in mind. It covers the denominator behind a rate, what a category was originally built to decide and what it flattens, the difference between a finding written about a community and the same finding written about the process that reached them, and a six-line check you can run before a number becomes a decision. Nothing here is a score, and nothing you write in a reflection is collected.

Responsible data and evaluation

One page for reading, writing and reporting a division number

A number is a record of what we counted. Who never reached the count is the question the number cannot answer and you can.

Keep the six questions in view while you request a figure, read a report, write a finding, or decide what a chart means before it travels.

Ask what the number is a share of before you ask whether it went up.
Every category was built for a purpose; say what it records and what it was built to decide.
A blank field is information about the conversation, not a value you may fill in.
Every difference has a process sentence and a people sentence. Write the process one first.
Unknown, written down, is the most useful answer a check produces.

Fill in the six lines for one report you produce or rely on, and run them once with someone who did not build the number.

Before you accept a figure

Ask what it is a share of, in one sentence.
Ask what happens to a case that leaves before the measure starts.
Ask who could never appear in it: never referred, never got through a first contact, recorded somewhere else.
Treat no complaints and no requests as findings about the route, not about the experience.

Before you break a number apart

Find the form or rule that defined each category, and who filled it in.
Say what is grouped inside each value, and how large the residual box is.
Count the blanks, and say what the report currently does with them.
Where small numbers must be suppressed, say so, and say what you did instead for those groups.

Before you write the finding

Write both sentences, the one about the people and the one about the process, and lead with the process.
State the comparison, and why that reference point rather than another.
Name what the data cannot separate, and leave the causal sentence out.
Put what the division already knows about that step in the same paragraph.

Before it travels

Put what the number counts on the chart itself, not in a footnote.
Share findings about a community with compensated advisors from it, and answer what they say in writing.
Say what will change and by when, or say plainly that nothing changes yet and when it will be looked at again.
Record unknowns as unknown, and name who could answer them.

Who this helps

Data, research and evaluation staff, who hold most of these questions already and often lack a moment where asking them out loud is expected — a useful place to start.
Quality, compliance and performance staff, whose measures decide what the division notices and what it never sees.
Policy and program staff, who carry a finding into a requirement and then live with what the categories could not show.
Contracts, fiscal, grants and procurement staff, whose reporting requirements decide what data exists at all, and about whom.
Executive and senior leaders, who choose which figure opens a briefing and whether unknown is a respectable answer — a useful place to start.

Related modules

Quality, complaints, and incident learning
Equitable policy and procedure governance
Co-design and compensated partnership
Resource allocation and equity
Leading a reflective DSD team

Participation and course completion in this program do not count toward DHS-required training credits unless management, a director, or DHS leadership expressly approves an exception.

Sources and further reading

Minnesota Department of Administration, Data Practices Office
The state's guidance on the Minnesota Government Data Practices Act, including how data about individuals is classified and what summary data is — the rules behind suppression and release decisions referenced in this module.
Minnesota State Demographic Center
State population estimates and demographic analysis, useful when an administrative count needs a population denominator it cannot supply on its own.
Washington Group on Disability Statistics
Internationally used question sets and guidance on how disability is defined and measured in data collection, including what short question sets capture and what they miss.
American Evaluation Association, Guiding Principles for Evaluators
Professional principles covering systematic inquiry, competence, integrity, respect for people, and responsibility for the general and public good — the reference behind this module's reporting practices.
U.S. Department of Health and Human Services, National CLAS Standards
National standards for culturally and linguistically appropriate services, including the standards on collecting and maintaining demographic data, conducting assessments, and partnering with communities.
Minnesota Council on Disability
Minnesota's advisory council on disability policy, access and rights, and a starting point for disability-led perspectives on how findings about disabled Minnesotans are framed.
PlainLanguage.gov, Federal plain language guidelines
Guidance on audience, organization and useful headings, applied here to writing a finding that can be read correctly by people outside the unit that produced it.
Minnesota Framework for Universal Multicultural Instructional Design
The design reference for this curriculum: multiple ways to engage, culturally responsive materials and accessible instruction.

All research and sources across the program

Opening your page…

You can still use the navigation while it loads.

Responsible Data and Evaluation

40–45 minutes · 4 lessons · Voluntary learning

Course outline

Who reaches the count

Division numbers describe the people who found the door, got through it and were recorded correctly. Everything that happens before the record is a filter, and the filters do not operate at random.

11 minutes

What the categories were built for

Every field in a record was defined by someone, for a reason, and the reason was usually operational. Trouble starts when a category built for payment or eligibility is asked a question about people's lives.

11 minutes

From a finding to a story

The same gap can be written as a fact about a community or a fact about the process that reached them. Both can be true. Only one of them names something the division can change.

10 minutes

Your data-interpretation checklist

Six questions to run before a number becomes a decision, short enough that you will actually use them, and written down so they do not depend on who happens to be in the room.

10 minutes

Sources and further reading

Minnesota Department of Administration, Data Practices Office

The state's guidance on the Minnesota Government Data Practices Act, including how data about individuals is classified and what summary data is — the rules behind suppression and release decisions referenced in this module.

Minnesota State Demographic Center

State population estimates and demographic analysis, useful when an administrative count needs a population denominator it cannot supply on its own.

Washington Group on Disability Statistics

Internationally used question sets and guidance on how disability is defined and measured in data collection, including what short question sets capture and what they miss.

American Evaluation Association, Guiding Principles for Evaluators

Professional principles covering systematic inquiry, competence, integrity, respect for people, and responsibility for the general and public good — the reference behind this module's reporting practices.

U.S. Department of Health and Human Services, National CLAS Standards

National standards for culturally and linguistically appropriate services, including the standards on collecting and maintaining demographic data, conducting assessments, and partnering with communities.

Minnesota Council on Disability

Minnesota's advisory council on disability policy, access and rights, and a starting point for disability-led perspectives on how findings about disabled Minnesotans are framed.

PlainLanguage.gov, Federal plain language guidelines

Guidance on audience, organization and useful headings, applied here to writing a finding that can be read correctly by people outside the unit that produced it.

Minnesota Framework for Universal Multicultural Instructional Design

The design reference for this curriculum: multiple ways to engage, culturally responsive materials and accessible instruction.