Phone Survey Response Rates: AAPOR Formulas
How to calculate phone survey response, cooperation, refusal and contact rates with AAPOR formulas, and how to map dialer dispositions to AAPOR codes.
Kurze Antwort
AAPOR's Standard Definitions (10th edition, 2023) define how survey outcomes are coded and how response rates are calculated. RR1, the minimum response rate, is complete interviews divided by all interviews, non-interviews and unknown-eligibility cases. Cooperation, refusal and contact rates use the same codes with different bases. To apply them, map each dialer disposition to a final AAPOR code, then calculate.
If you run phone surveys, someone will ask for the response rate. The trouble is that "response rate" has many definitions, and the official one is written for survey researchers. This guide walks through the American Association for Public Opinion Research (AAPOR) standard in plain English, shows a worked example and maps everyday dialer dispositions to AAPOR outcome codes. The tenth edition, published in 2023, is the current one at the time of writing (AAPOR, checked 2026-11-19).
One caution up front. AAPOR's codes depend on your sample design, and the standard says researchers must describe their design and justify their assumptions about eligibility. The table in this guide is a starting point for a list sample of named people, not a substitute for the standard.
What are the AAPOR outcome codes?
Every case in your sample ends in one final outcome. The formulas use these letters:
| Letter | Bedeutung | AAPOR code |
|---|---|---|
| I | Complete interview | 1.1 |
| P | Partial interview | 1.2 |
| R | Refusal and break-off | 2.10 |
| NC | Non-contact | 2.20 |
| O | Other eligible non-interview | 2.30 and 2.90 |
| UH | Unknown if household or occupied housing unit | 3.10 |
| UR | Unknown if the sampled unit is eligible, or the household has an eligible respondent | 3.20 |
| UO | Unknown, other | 3.90 |
| e | Estimated proportion of unknown-eligibility cases that are eligible | An estimate, not a code |
There is also a "not eligible" group (codes 4.0), such as a person screened out of the sample. Those cases are left out of every formula below. The key idea: eligible cases are known to belong in the sample, unknown cases might, and not eligible cases do not.
How do you calculate response rate RR1?
RR1 = I / ((I + P) + (R + NC + O) + (UH + UR + UO))
In words: RR1 is the number of complete interviews divided by the interviews (complete plus partial), plus the non-interviews (refusals and break-offs, non-contacts, others), plus every case of unknown eligibility. AAPOR calls it the minimum response rate, a lower bound, because it assumes every unknown case was eligible.
There are six response rates in the standard. The others change one of two things: whether partials count as respondents, and how unknown cases are treated.
| Rate | Numerator | Unknown cases in the base |
|---|---|---|
| RR1 | I | All counted (UH + UR + UO) |
| RR2 | I + P | All counted |
| RR3 | I | Only the estimated eligible share: e x (UH + UR + UO) |
| RR4 | I + P | Estimated share, as in RR3 |
| RR5 | I | None (assumes e = 0, or there are no unknowns) |
| RR6 | I + P | None (the maximum response rate) |
AAPOR says RR3 can be considered the most common AAPOR response rate in reporting, and says RR5 and RR6 are only appropriate when it is valid to assume that none of the unknown cases are eligible, or when there are none.
A worked example, with illustrative counts
These counts are made up to show the arithmetic. They are not results from any survey. Suppose a list of 1,000 sampled phone numbers ends like this:
| Ergebnis | Cases |
|---|---|
| I (complete) | 300 |
| P (partial) | 40 |
| R (refusal and break-off) | 250 |
| NC (non-contact) | 200 |
| O (other) | 10 |
| UH + UR + UO (unknown eligibility) | 200 |
The base for RR1 is (300 + 40) + (250 + 200 + 10) + 200 = 340 + 460 + 200 = 1,000. So RR1 = 300 / 1,000 = 30.0%. RR2 = 340 / 1,000 = 34.0%. If you estimate that half of the unknown cases are eligible (e = 0.5, and you can justify it), the base for RR3 is 340 + 460 + 100 = 900, so RR3 = 300 / 900 = 33.3%.
Cooperation, refusal and contact rates
These use the same letters, with different bases. Each answers a separate question.
| Rate | Formel | Question it answers | Example |
|---|---|---|---|
| COOP1 | I / ((I + P) + R + O) | Of eligible people we reached, how many completed? | 300 / 600 = 50.0% |
| REF1 | R / ((I + P) + (R + NC + O) + (UH + UR + UO)) | What share of the sample refused or broke off? | 250 / 1,000 = 25.0% |
| CON1 | ((I + P) + R + O) / ((I + P) + (R + NC + O) + (UH + UR + UO)) | What share of the sample did we reach? | 600 / 1,000 = 60.0% |
Read them together. A low response rate with a high cooperation rate means you are not reaching people: the problem is contact, which is a matter of list quality, calling times and retries. A low cooperation rate means people are reached but decline: that points to the introduction, the length or the topic. AAPOR also defines COOP2 to COOP4, REF2 and REF3, and CON2 and CON3, which vary the base the same way the response rates do. Our glossary entry on Kontaktrate covers the dialer-side idea of the same name, which is counted differently.
How do dialer dispositions map to AAPOR codes?
Your dialer records a disposition for each attempt. AAPOR wants one final code per case. The standard treats things like "call back" and "maximum call limit met" as temporary codes that must be replaced with a final code when fieldwork ends. The table below is a starting map for a list sample of named people. Confirm every row against the tables in the standard for your design.
| Dialer disposition | Usual AAPOR outcome | Note |
|---|---|---|
| Completed survey | I (1.1) | Define in advance how much must be answered to count as complete |
| Teilweise | P (1.2) | Needs a written rule for enough answered items |
| Refused | R (2.10) | A refusal before eligibility is confirmed may be unknown eligibility instead |
| Hung up mid-survey | R as break-off (2.12), or P | Depends on how far they got against your partial rule |
| Booked a call-back but never took it | R (implicit refusal, 2.1132) | The standard lists an appointment set but not kept, and recommends coding such cases as a refusal unless there is evidence otherwise |
| Answering machine, message confirms it is the named person's number | NC (2.22) | Only where no extra screening is needed. Sub-codes: 2.222 message left, 2.221 no message left |
| Answering machine, greeting does not say whose number it is | Unknown eligibility | Standard code is 3.2153 where screening is required |
| No answer or busy, every attempt | Unknown eligibility | No-answer and busy have their own sub-codes |
| Disconnected or non-working number | Unknown eligibility (3.216) | When in doubt, the standard says to presume a case eligible or possibly eligible, not ineligible |
| Wrong number, person's eligibility already confirmed | O (2.37) | The number dialled is incorrect for the named person |
| Screened out (not in the target group) | Not eligible (4.10) | Left out of the formulas |
| Language or technical barrier, or respondent unable to take part | O (2.30: language 2.33, unable 2.32) | Eligible, but no interview possible |
The standard also gives a precedence rule. Once there has been a refusal, the case stays a refusal unless it was converted into an interview or you later learn it was ineligible. Repeated no-answers after a refusal do not turn it into a non-contact. Build the rule into your export logic so the final code is not simply whatever the last attempt happened to be.
Why AMD and retry settings change your rates
Two dialer settings move cases between categories, so they belong in your methods note.
- Answering-machine detection. If your system hangs up on a detected machine without leaving a message, those cases fall into non-contact or unknown eligibility, depending on whether the greeting identified the person. Dropping a message changes the sub-code. See the glossary entry on Anrufbeantwortererkennung for how detection works.
- Attempt caps. A cap of three attempts can leave more unreached cases than a cap of ten. Those cases go to non-contact or unknown eligibility, which lowers RR1. Spreading attempts across days and times changes who is reached. Our guide to retry strategy by disposition shows how to set caps by outcome.
Neither setting is wrong, but changing them mid-fieldwork changes what your rates mean. Note the policy and the date of any change.
Reporting your rates honestly
AAPOR's guidance is explicit on a few points. State which rate you used, for instance "Response Rate 3 was 33.3%", and make a table of final disposition codes for all cases available. When you use an eligibility estimate e, state how you derived it, and do not choose a value to raise the response rate. A good methods paragraph includes:
- The formula used (RR1, RR3 and so on) and the AAPOR edition.
- The sample source and size, and how eligibility was determined.
- The estimate e and its basis, if applicable.
- The attempt policy: maximum attempts, calling windows and answering-machine handling.
- Your rule for a complete interview and a partial one.
- The table of final outcomes by code.
How DialerBee supports survey calling
Der survey dialer page lists CATI outbound interviewing, panel recruitment, quota-based sample management, respondent callback scheduling, call-recording controls and surveys across 11 languages. It also lists APIs and webhooks for dispositions, so that outcomes flow back to panel systems. The Umfragen und Kundenerfahrung feature describes outbound survey campaigns with pacing and quotas, and says reporting shows each figure with its denominator stated. The Berichte page covers dispositions and exports. DialerBee does not state that it computes AAPOR rates for you, so export final dispositions, map them as above, and calculate in your own tool. For feedback programmes, read our guide to CSAT and NPS surveys over WhatsApp and voice. The glossary explains Anrufstatus.
Recording and consent settings are configured by the research organisation to match its own policies and local rules. DialerBee provides compliance-supporting controls and does not guarantee compliance.
Häufig gestellte Fragen
What is the AAPOR formula for response rate 1 (RR1)?
RR1 = I / ((I + P) + (R + NC + O) + (UH + UR + UO)). I is complete interviews, P partial interviews, R refusals and break-offs, NC non-contacts, O other eligible non-interviews, and UH, UR and UO are the three kinds of unknown eligibility. RR1 is the minimum response rate: it counts every case of unknown eligibility as if it were eligible.
What is the difference between a response rate and a cooperation rate?
A response rate divides completes by all eligible and possibly eligible cases in the sample. A cooperation rate divides completes by the cases where an eligible person was actually contacted. COOP1 = I / ((I + P) + R + O). Because the base is smaller, the cooperation rate is at least as high as the response rate, and it tells you how many people who were reached agreed to take part.
What is the contact rate and how is it calculated?
AAPOR's CON1 = ((I + P) + R + O) / ((I + P) + (R + NC + O) + (UH + UR + UO)). It is the share of the whole sample in which someone responsible was reached. It does not count whether they agreed. CON1 treats every unknown-eligibility case as eligible; CON2 uses the estimate e; CON3 counts only known eligible cases.
Which AAPOR response rate should I report?
AAPOR says to report exactly which rate you used, for example that Response Rate 2 was a stated figure, and to make a table of final disposition codes available. AAPOR says RR3 can be considered the most common one in reporting. Whichever you choose, state how you estimated e, the proportion of unknown-eligibility cases that are eligible, and do not pick a value to raise the rate.
How do answering machines count?
It depends on what the message tells you. In the AAPOR list-sample tables, a phone answering device that confirms it is the number for the named sample member is a non-contact (2.22), a code used only when no extra eligibility screening is needed. An answering device that does not make clear whose number it is falls under unknown eligibility. Whether you left a message is recorded as a sub-code. Check the tables for your sample design before you map anything.
Can I use these formulas for a short customer satisfaction call?
You can calculate the same ratios for any list-based phone survey, but AAPOR's standard is written for survey research and its outcome codes assume a defined sample. For a CSAT or NPS programme, the main discipline carries over: count every sampled case, assign it a final outcome, and state the denominator you used.
Quellen
- AAPOR, Standard Definitions: current edition is the 10th, 2023. aapor.org, Standard Definitions (geprüft am 19.11.2026)
- The American Association for Public Opinion Research. 2023 Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 10th edition. aapor.org, Standard Definitions 10th edition (PDF) (geprüft am 19.11.2026)
Weiterführende Lektüre
- Survey Dialer for Market Research
- Umfragen und Kundenerfahrung
- CSAT and NPS Surveys Over WhatsApp and Voice
- Was nennt man Disposition?
- Was ist die Kontaktrate?
- Was ist Anrufbeantwortererkennung?
- Call Retry Strategy by Disposition
- Berichte und Insights
Note: This article summarises the AAPOR Standard Definitions, 10th edition (2023), as published on aapor.org and checked on 2026-11-19. It is not a substitute for the standard, and the official text always prevails. Counts in the worked example are illustrative. DialerBee does not provide legal advice or guarantee regulatory compliance.
Verwandte Artikel
How to Reduce No-Shows With Reminder Calls
10 Minuten Lesezeit
IndustrieLeitfaden für ausgehende Anrufe von Autohändlern
12 Minuten Lesezeit
FinanzdienstleistungenSAMA Rules for Bank Marketing Calls
11 Minuten Lesezeit
EinhaltungSpain's 400 Prefix for Sales Calls in 2026
11 Minuten Lesezeit
Bereit, DialerBee in Aktion zu sehen?
Fordern Sie eine 15-minütige Live-Demo an oder starten Sie eine 14-tägige kostenlose Testphase – ohne Folien, ohne Verpflichtung. Das Telefonieren beginnt, sobald Ihr eigener SIP-Trunk oder Ihre TK-Anlage verbunden und freigegeben ist.
14-tägige kostenlose Testphase · kostenloses KI-Guthaben · keine Kreditkarte · 11 Sprachen · BYOC · Kontrollen zur Unterstützung der Compliance