Voice spelling is a special case for alphanumeric input (see Alphanumeric Input).

Due to the high acoustic confusability of the spoken letters of the alphabet, it's best to avoid dialog steps that require voice spelling (Lewis, 2011; Rolandi, 2006; Wright et al., 2002).

Some voice spelling modules accept words from the phonetic alphabet (e.g., alpha, bravo, charlie, delta, echo, foxtrot, golf, hotel, india, juliet, kilo, lima, mike, november, oscar, papa, quebec, romeo, sierra, tango, uniform, victor, whiskey, x-ray, yankee, and zulu) – but you can't count on most callers knowing the entire phonetic alphabet (for a list of alternative words to consider when writing American English voice spelling alphabets, see Lewis, 2005). Also, if words other than those for letters are active, be sure to conduct grammar testing to make sure there are no unintended confusions (e.g., Lewis & Commarford, 2003, reported confusion between “juliet” and “period” when voice spelling short sentences).

Lewis, J. R. (2005). Frequency distributions for names and unconstrained words associated with the letters of the English alphabet. In Proceedings of HCI International 2005: Posters (pp. 1–5). St. Louis, MO: Mira Digital Publication. Available at http://drjim.0catch.com/hcii05-368-wordfrequency.pdf.

Lewis, J. R. (2011). Practical speech user interface design. Boca Raton, FL: CRC Press, Taylor & Francis Group.

Lewis, J. R., & Commarford, P. M. (2003). Developing a voice-spelling alphabet for PDAs. IBM Systems Journal, 42(4), 624–638.

Rolandi, W. (2006). The alpha bail. Speech Technology, 11(1), 56.

Wright, L. E., Hartley, M. W., & Lewis, J. R. (2002). Conditional probabilities for IBM Voice Browser 2.0 alpha and alphanumeric recognition (Tech. Rep. 29.3498) West Palm Beach, FL: IBM.