Church survey questions, and the one answer you cannot count
A 15-question survey for a congregation, ready to send. The three questions at the front tell you who answered, because a church has no membership list to check them against. Answer it here first, then send your own.
- 15questions
- about 4 minto complete
- 3asked first, to see who replied
- No nameand no email address
The one answer in a church survey you cannot count
Almost everything a congregation asks can be checked against something it already knows. One question cannot, and it is the one every church asks first.
There is no list to check the answers against. A club posts a survey over a roster and can set who replied against who is on it. Nobody signs in on a Sunday, so the survey's own attendance question is the measurement rather than a check on it, and it is the one question here with a known direction of error.
Asked by a person, people report attending more often than they do.[1] Pew Research Center put the same question to US adults both ways. With a live interviewer on the telephone, 43 per cent said they attend religious services monthly or more; self-administered, 33 per cent did, and weekly attendance was 31 per cent against 26.[1] Pew puts the gap down to social desirability, and to some people answering a question about how often they attend as a question about who they are.[1]
So your own number will read low, and that is the survey working. A link sent to a congregation is self-administered, the mode Pew found gives the more candid answer.[1] Say that before anyone sees the results, or the first conversation is about whether the survey is broken. The direction transfers to you; the ten-point gap does not.
Which is why question 1 asks about four weekends. A bounded stretch has an answer somebody can remember; "how often do you typically attend" invites the identity reading instead.[1] This line of work starts with Hadaway, Marler and Chaves in 1993, in "What the Polls Don't Show: A Closer Look at U.S. Church Attendance".[2]
Whatever the survey says, it is not a head count. If you want a count, count.
The 15 church survey questions
Every question, the answer format, and what a weak answer is telling you. Copy one block or the whole set.
Who is answering
Asked first and required. A half-finished response still says who it came from.
- Of the last four weekends, how many did you attend a service here?A count of a stretch you can remember, not a typical week. Read every other answer split by it.
- How long have you been connected with this church?If nobody in the first two bands answered, you have heard from the settled half only.
- Are you currently part of a group, team or volunteer role here?The people who picked "I used to be" are the shortest useful list on the page.
The services
Two items about the thing people are there for. Neither asks anyone to rate their own faith.
- Overall, how satisfied are you with the worship services?One overall number to repeat next year with the same three questions in front of it.
- The music supports meaningful worship.Music splits by age more than anything else here, so read it with question 15 if you asked it.
Belonging and care
Being welcomed at the door and finding a way in are different things, so they are two questions.
- I feel welcome here.Compare the newest tenure band against the longest. A gap here is a joining problem.
- How easy is it to find opportunities to get connected or serve?Low scores from people who answered "No" at question 3 are the ones to act on.
- How satisfied are you with the support available when someone needs help?Many people will have no experience to draw on. Report the count who answered, not just the average.
Communication and the building
Four things a congregation can change this month, including the one money question, which is about being told.
- Church communications keep me informed about what is happening here.Split by question 1. People who come once a month are the ones an announcement never reaches.
- I understand how money given to this church is used.This asks whether people are told, not what they give. Nothing here asks for an amount.
- The building here is well looked after.A low score here is a maintenance list, and it is the cheapest thing on this page to fix.
- When someone raises a concern here, something happens about it.The one item on this page that a survey with no visible follow-up makes worse. Read it again next year.
The close
One single-tap question, one open box and one optional item, so it does not end on a blank field.
- How likely would you be to invite someone you know to come with you?Five worded options rather than a 0 to 10 line, so nobody turns it into a net promoter score.
- What is one thing we could improve for the coming year?Sort these by question 1 before you count them. The occasional attenders ask for different things.
- Which age band are you in?Three broad bands, and the only thing asked about the person. Cut it in a small congregation.
Two edits before you send. Cut question 5 if music is not part of what you do, and put the names your congregation actually uses into question 3.
Reading a church survey when seventy people answered
Almost every piece of church survey advice online is written for a congregation several times the size of yours. Two things change when it is small.
The typical congregation and the typical churchgoer are not in the same room. In the 2018-19 National Congregations Study, the median US congregation had about 70 people regularly taking part; the congregation attended by the median US churchgoer had about 363.[3] Both are true at once, and the bigger one is what most advice is written for. Faith Communities Today, surveying 15,278 US congregations, put median weekly worship attendance at 65 in 2020 against 137 in 2000.[4]
At that size a demographic grid names people. Take the same study's medians: 50 of those 70 are adults, 6 in 10 are women, 1 in 5 is under 35, and 3 in 10 held a leadership role last year.[3] Multiply those through and "a woman under 35 who serves on a team" is about two people. Ask for an age band, a gender, a tenure band and a role in one survey and you have not collected demographics, you have collected names. Hence one optional age band and nothing else about the person.
Read every other answer split by question 1. The people who answered "all four" and the people who answered "one" are two congregations sharing a building, and an average across them describes neither. Print both columns. Question 2 gives you a second split, newest against longest connected, which shows whether the way in works.
| What the split shows | What it means | What to do about it |
|---|---|---|
| Nearly everyone answered "all four" at question 1 | You have heard from the people who are always here. | The handful who answered "one" or "none" are the finding. Read them first, however few. |
| The newest tenure band scores lowest on question 6 | Being welcomed and finding a way in are coming apart. | That is a joining problem, and it is fixable without a budget. |
| Question 9 is low only among occasional attenders | Announcements reach the people already in the room. | Whatever channel you rely on, it is not the one they use. |
| A group returned four answers and a striking number | Four answers is four people, not a trend. | Print the count beside the percentage and call the line directional. |
Report counts, not just averages. Hartford Institute for Religion Research, which has run congregational inventories since 2004, will not give churches cross-tabulations at all: reading them needs significance testing, and committees take a chance difference for a finding.[5] Short of that, put the number of people behind every percentage, and where a group is small say so on the line.
If question 11 comes back about the hall, the car park or the heating, that is a building being judged, and the building and facilities questions go into it properly.
What to change before you send it to your congregation
One decision, two edits, and the note to put in front of people.
Decide who can open the answers, then say so. The template asks for no name and no email address, but somebody holds the login and can read every response, and at seventy people that is somebody who knows every respondent by name.[3] Put their name in the invitation, and say what will be reported, at what level of grouping, and what will not be.
Cut question 15 if your congregation is under about forty people. One optional item about the person is all a small group can carry, and where to place an optional age question matters more than which bands you choose.
Question 10 is about being told, not about giving. It asks whether people understand how money given is used, which is a communication question. When the subject is the giving itself, why people give and what would bring them back, the donor survey questions are the right set instead.
Three things this template is not. Not a pastor search instrument: that is a hiring decision with a search committee, and a different obligation on whoever answers. Not a visitor survey: it assumes the person answering already comes. And if what you want judged is one occasion rather than the year around it, ask for feedback on a single event instead.
A youth group, a food bank or a course has an aim of its own, and the program feedback questions measure it against that aim.
If the people you are asking hold a paid membership with a renewal date rather than turning up on a Sunday, the member survey template is built for that list.
All fifteen questions are already loaded in the questionnaire creator, so the only editing left is your own wording.
Do not ask what you will not answer. Question 14 invites somebody to name one thing to change, and question 12 asks whether raising something here leads anywhere. If nothing visible follows, you have answered question 12 for them.
What to put in front of people before question one
Before you start This survey asks how things are working here: the services, whether it is easy to find a way in, whether communications reach you, and what you would change. The first three questions ask how often you have been here recently, how long you have been connected, and whether you are part of a group or team. They are there so we can tell whether the people who answered look like the people who come. Nothing here asks for your name or an email address. [NAME] can open the responses. We will report answers as groups, never as one person, and we will not quote anything that would identify who wrote it. Fifteen questions, about four minutes. It closes on [DATE], and we will say what we heard and what we are doing about it by [DATE].
Church survey FAQ
What questions should you ask in a church survey?
Three about who is answering, then the things a congregation can change: the services, finding a way in, whether communications reach people, the building, and whether a raised concern goes anywhere. Nothing asking somebody to rate their own faith, which turns a feedback survey into a test with a right answer.
How many questions should a church survey have, and how long should it take?
Fifteen, about four minutes. The published congregational instruments are a different product: Hartford Institute's sample states approximately 30 to 45 minutes.[6] The free question lists online run past a hundred, which is a menu to choose from rather than a survey anyone sends.
How do you ask about church attendance without getting an inflated number?
Ask about a bounded stretch. Question 1 asks how many of the last four weekends somebody was here, which has an answer they can remember, rather than how often they typically attend, which Pew found some people answer as a question about who they are.[1]
Should a church survey be anonymous?
Nobody is asked to identify themselves. Whoever owns the survey can still open every response, and in a congregation that is somebody who knows the people answering, so name them in the invitation and say results go out as groups.
Should we pay for a congregational assessment instead?
They are different products. Hartford Institute charges $300 for a congregational inventory and publishes comparison data from 46,947 respondents across 323 mostly mainline churches, at an average response rate of roughly 30 to 35 per cent of total membership.[6] What the money buys is the report, the comparison and somebody who has seen the other churches. This is fifteen questions you can send this week.
Methods and sources
What this template is based on
No published congregational instrument was reproduced, because there is no free reprintable one. The design is three participation questions asked first, two on the services, three on belonging and care, four on communication, money and the building, one close, one open box and one optional age band. Question 1 asks about a bounded recent stretch rather than a typical week because Pew found that phrasing invites an answer about identity.[1] The demographics are cut rather than expanded because of the arithmetic of a 70-person congregation.[3]
Question licensing
Hartford Institute's Church Planning Inventory, Parish Profile Inventory and Pastoral Search Inventory are downloadable but are a $300 commercial service, so they are named and cited here and no item from them is reprinted.[6] The Sense of Community Index is published free of charge but may not be modified and requires a completed request form, which is why no belonging scale is reproduced. Holy Cow Consulting's Congregation Assessment Tool, Barna's assessments and Natural Church Development are commercial and unpublished, and none of their benchmark claims appears here.
Limits and disclosure
The Pew figures are US adults answering a national survey, not members of one congregation answering their own church, so what transfers is the mechanism and its direction rather than the percentages. The congregation medians are US and are medians, so half of all congregations sit above them. The Hartford response rate is one organisation's published experience over 2004 to 2014, from churches that chose to buy a $300 instrument and were mainline-heavy, and it is a rate of total membership; it is not a target to hit.[6] Hadaway, Marler and Chaves is named as the origin of this line of work and no figure from it is quoted, because the copy available to us is a scan we could not read. The survey carries no score: it produces counts by group, and there is nothing here to add up.
This template takes no position on how a congregation should be governed or what it should believe. Nothing in it asks anyone to rate their own faith. SuperSurvey builds the editor it opens in, and nothing here depends on using it.
References
- Pew Research Center. Measuring Religion in Pew Research Center's American Trends Panel. Washington, DC: Pew Research Center, 14 January 2021. pewresearch.org
- Hadaway, C. Kirk, Penny Long Marler and Mark Chaves. "What the Polls Don't Show: A Closer Look at U.S. Church Attendance." American Sociological Review 58, no. 6 (1993): 741-752. Named here as the origin of this line of research; no figure from it is quoted on this page.
- Chaves, Mark, Joseph Roso, Anna Holleman and Mary Hawkins. National Congregations Study: Waves I-IV Summary Tables. Durham, NC: Duke University Department of Sociology, 2021. Wave IV, 1,262 congregations, 2018-19. sites.duke.edu
- Faith Communities Today. Twenty Years of Congregational Change: The 2020 Faith Communities Today Overview. Hartford, CT: Hartford Institute for Religion Research, 2021. faithcommunitiestoday.org
- Hartford Institute for Religion Research. "Survey Your Congregation." Hartford International University. hirr.hartfordinternational.edu
- Hartford Institute for Religion Research. Church Planning Inventory Sample and Congregational Member Survey Summary: 46947 Total Respondents. Hartford International University, 2004-2014. Inventory sample, comparison report
Send the church survey, and read it by who answered
Fifteen questions, about four minutes, and three at the front that tell you whether the answers describe your congregation or only the part that never misses a week.
Use this template