Validation

=Validation Protocol= ==

The scale developed here aims to identify existing hearing loss, and risk factors for potential hearing loss, in order to establish a screening process by which members of the public can be referred on for further testing and/or treatment. The results of this test may have direct implications on the health of individuals, and as such, a validation protocol must be observed, in order to confirm that inferences about individuals can reliably be made from this scale. In order to establish content validity, the items included in this questionnaire will need to be representative of the constructs that we are aiming to measure. Presently we are trying to measure five main risk factors: To that end, the items in this test need to reflect all aspects of these factors, or at least measure all variables which could explain any variance in them. For example, measuring risk factors for potential hearing loss (for example) needs to include all possible items which relate to this. If we focus solely on items which relate to risk factors from workplace related noise, this would ignore other potential activities which may contribute to hearing loss. In this regard, it will also be important to collect age group data, to control for normative hearing loss related to age. Establishing criterion related validity is particularly important in reference to the measurement of existing hearing loss. Concurrent validity could be established by measure the test against objective measures of hearing loss (a full audio assessment). Predictive validity would be important to establish in measures of risk factors for potential hearing loss. It seems important to confirm that a high score on risk factors for hearing loss would correlate with objective hearing loss in future (if aforementioned risk activities are not controlled).
 * 1) Can we state that we plan to continue validation of the study after ethics approval?
 * 2) This is relevant for all types of validity
 * 3) Somewhat necessary given that a study cannot be valid if it is not reliable, impossible to establish reliability without first collecting data
 * 4) The judging issue in content validity
 * 5) Can we say that we plan to do this?
 * 6) Do we mention that it isn’t possible?
 * 7) Do we mention it at all?
 * Existing hearing loss
 * Work related hearing damage
 * Hobby related hearing damage
 * Attitudes to hearing loss
 * Demographic data, age and gender

Construct validity will most likely be the most applicable in the initial stages of this assignment. This will involve substantial literature review to determine both convergent and discriminant validity. This sort of validation lends itself in particular to attitudes towards hearing loss, for instance, it would need to be established that a poor attitude to hearing protection would correlate reasonably with other psychological measures such as problems with authority. On the other hand, we would not expect it to be correlated with unrelated measures.
 * (?) Predictive validity would follow the formula y=mx+c, where y = need for further hearing protection/intervention, m = risk factor score, x = time, c = existing hearing loss
 * A validation protocol would need to ensure that individuals can be reliably placed in this context in order to actually show that a certain level of hearing damage will be reached at a certain point if no intervention is made. E.g. If person A has 20% hearing loss and has a risk factor score of 22, can we predict that person A will have 40% hearing loss in one year

Gregory, R. J. (2011), Psychological Testing: History, Principles, and Applications (6th ed.), //Pearson Education//, Boston

Round 2

Guys: The following is the details of an article that constructed and validated a measure of existing hearing loss (pdf wasn't available).

The Development and Cross-Validation of a Self-Report Inventory to Assess Pure-Tone Threshold Hearing Sensitivity.

Coren, StanleyHakstian, A. Ralph

Journal of Speech & Hearing Research; Aug92, Vol. 35 Issue 4, p921, 8p

[|http://web.ebscohost.com/ehost/detail?sid=5d12c1c7-01d7-486e-8343-6f7a0a5c3f8c%40sessionmgr115&vid=1&hid=105&bdata=JnNpdGU9ZWhvc3QtbGl2ZSZzY29wZT1zaXRl#db=ufh&AN=4435267]

=Work-related, noise-induced hearing loss: evaluation including evoked potential audiometry.= [|Barrs DM], [|Althoff LK], [|Krueger WW], [|Olsson JE].

http://www.ncbi.nlm.nih.gov/pubmed/8108153