A parity bit is an extra bit added to a group of data bits so the total count of 1s follows a fixed rule — always even (even parity) or always odd (odd parity). When received data breaks that rule, the receiver knows at least one bit was flipped in transit.
Why do data transmissions need error detection?
Every time data travels — across a network cable, through the air as a Wi-Fi signal, or even between RAM and the CPU — there is a small chance that electrical interference, radio noise, or hardware defects will flip a bit from 0 to 1 or vice versa. Without error detection, the receiver has no way of knowing that the data it received differs from the data that was sent.
Professor Turing's analogy: imagine sending a text message that says "MEET AT NOON" and one letter is changed by a noisy channel to "MEET AT MOON". If you had a rule that every message must contain exactly 10 letters, you would still receive 10 letters and would not notice. A better rule — one that changes when any single character changes — gives you a chance to detect the corruption.
How does even parity work?
In even parity, the sender adds a parity bit so the total number of 1s in the block (including the parity bit) is always even.
Worked example 1 — transmitting the byte 1011001:
| Data bits | Count of 1s | Even parity requires | Parity bit |
|---|---|---|---|
1 0 1 1 0 0 1 |
4 (already even) | Total must be even | 0 |
Full transmission: 1 0 1 1 0 0 1 **0** (8 bits, four 1s)
Worked example 2 — transmitting 1101001:
| Data bits | Count of 1s | Even parity requires | Parity bit |
|---|---|---|---|
1 1 0 1 0 0 1 |
4 (already even) | Total must be even | 0 |
Wait — recounting: 1+1+0+1+0+0+1 = 4. Parity bit = 0. Total 1s = 4. ✓
Worked example 3 — transmitting 1101011:
| Data bits | Count of 1s | Even parity requires | Parity bit |
|---|---|---|---|
1 1 0 1 0 1 1 |
5 (odd) | Total must be even | 1 |
Full transmission: 1 1 0 1 0 1 1 **1** (8 bits, six 1s)
How does a parity check detect an error?
The receiver counts the 1s in the received block. If the count violates the agreed parity rule, an error is flagged.
Detection example: suppose the transmitted block was 10110010 (even parity, four 1s) but noise flipped the third bit:
- Transmitted:
1 0 **1** 1 0 0 1 0— four 1s ✓ - Received:
1 0 **0** 1 0 0 1 0— three 1s ✗ PARITY ERROR
The receiver immediately knows the data is corrupt and can request a retransmission.
In odd parity, the same process applies but the target is an odd total. Some systems use odd parity because an all-zeros block would have zero 1s — an even number — which would incorrectly pass even-parity checks on a dead link.
What is a checksum and how is it different from a parity bit?
A checksum is a calculated value derived from a larger block of data. The sender computes the checksum and appends it to the data; the receiver recalculates the checksum from the received data and compares the two.
| Feature | Parity bit | Checksum |
|---|---|---|
| Size | 1 bit per block | Typically 8, 16, or 32 bits |
| Errors detected | Single-bit errors only | Multi-bit errors in most cases |
| Errors corrected | None — can only detect | None (detection only) |
| Computation cost | Trivial (count the 1s) | Slightly higher |
| Used in | Simple hardware transmission | TCP/IP networking, file downloads |
A checksum is more powerful but requires more overhead. The Internet Protocol (IP) and TCP both carry checksums to protect packet headers and data.
What are the limitations of parity?
Parity has two important limitations:
-
It cannot detect even-numbered bit errors. If two bits are flipped, the parity count returns to its original state and the error is invisible. This is called a double-bit error slipping through.
-
It cannot correct errors — only detect them. When a parity error is found, the receiver can only request retransmission; it cannot determine which bit was flipped and fix it.
For situations where retransmission is impossible (e.g. data written to a memory chip where the write is gone), more advanced techniques such as Hamming codes are used. These add enough extra bits to locate the exact bit that changed and correct it automatically — but that level of detail is beyond GCSE.
Frequently asked questions
How do I calculate the parity bit in an exam question?
Count the number of 1s in the data bits. For even parity: if the count is already even, the parity bit is 0; if the count is odd, the parity bit is 1. For odd parity: if the count is already odd, the parity bit is 0; if even, the parity bit is 1. Always recount after adding the parity bit to verify the total is correct before writing your answer.
Why does a double-bit error go undetected by parity?
If two bits are flipped, one flip adds a 1 (or removes one) and the second flip does the opposite — the net change to the count of 1s is zero. The parity rule is still satisfied, so the receiver has no way to tell anything went wrong. This is the fundamental weakness of single-bit parity: it only catches errors that change the parity of the block.
Is parity still used in modern computers?
Yes — ECC (Error-Correcting Code) RAM uses a more advanced form of parity called Hamming parity to detect and correct single-bit errors and detect double-bit errors in memory. Server-grade hardware commonly uses ECC RAM. For network transmission, parity alone has been superseded by CRC (Cyclic Redundancy Check) and TCP checksums, which are far more powerful. Parity remains in the GCSE specification because it introduces the core principle of error detection elegantly.
What is the difference between error detection and error correction?
Error detection means discovering that the received data differs from what was sent; the receiver then requests retransmission. Error correction means the receiver has enough redundant information to determine exactly which bits are wrong and fix them without a retransmission. Detection requires fewer extra bits; correction requires more. At GCSE level, parity is covered as a detection mechanism — correction techniques appear at A-level.
For step-by-step parity walkthroughs and every other GCSE data representation topic, try Professor Turing at aitutors.me.