Words-in-Noise Scoring

The Words-in-Noise test, available in NIH Toolbox V2 and NIH Toolbox V3, measures the participant’s ability to recognize single words amid varying levels of background noise. Each ear is tested separately via headphones, starting at 24db, and going as low as 4db. The V3 Scores Export / V2 Assessment Scores shows the raw score (i.e., number of words named correctly) by ear.

The V3 Item Export / V2 Assessment Data shows the coded Response and Score for each signal-to-noise ratio (SNR). That is, rather than listing which words were named correctly, each SNR is coded as a bitwise score. To determine the number of words correct per SNR, you can use the table below to help calculate the bitwise Response/Score.

1. For each SNR (i.e., each row in the V3 Item Export / V2 Assessment Data file), determine which bitwise values add up to the value in the Score column. For example, if you see a Response/Score of 23 for ItemID AUWINLeftList2_24db, the participant must have answered the first, second, third, and fifth words correctly (1+2+4+16=23) at the 24db level in the left ear. As another example, a Response/Score of 13 means the participant answered the first, third, and fourth words correctly (1+4+8=13). 
2. Step 1 tells you which words were answered correctly, and therefore how many words were answered correctly per SNR. In the example above, the participant answered 4 words correctly in the 24db Left Ear SNR item.
3. To get the raw score per ear, sum the number of words correct per ear.

	Bitwise Value
First Word	1
Second Word	2
Third Word	4
Fourth Word	8
Fifth Word	16