The process helps determine whether behaviours are well defined, accurate and recorded consistently. This is also called program correction and emphasizes the importance of accurate data recording. Person C shows serious self-injury and therefore receives a thorough assessment of the behaviour by a qualified team. Following the results of the evaluation, a differential-strengthening procedure is required. The parents of Person C are instructed to perform the procedure and record the self-injury data before and after the treatment. Parents are diligent and reasonably accurate with data collection before treatment, but they leave a little after the start of treatment and forget to enter many instances of self-injury. At the next interdisciplinary professional team meeting, the team concludes that behavioural treatment based on parental data has been effective and that no changes are being made. In fact, data reliability tests showed that behavioural treatment was ineffective. The differential-strengthening procedure has not worked and other potentially effective behavioral methods (for example. B other variations of differential amplification) are left on the shelf. The reliability of the data and the integrity of treatment have important effects on clinical practice, as they can influence the ability of physicians to accurately assess the effectiveness of behavioural interventions. Reliability and integrity data also allow physicians to provide feedback to nurses and tailor interventions when needed. We explain why reliability and integrity measures are of the utmost importance in clinical work, discuss events that may reduce reliability or integrity, and provide several effective ways to collect data and calculate reliability and integrity measures.

Another method of reliability is used when there is a partial interval or a full interval. The partial interval refers to the interval score if the behavior occurs at any point in that interval. The full interval refers to the assessment of the interval when the behaviour occurs throughout the interval. Therefore, there is no “counting” of behaviour; the interval is simply considered a “deposit” or a “non-deposit.” In the case of interval recording, reliability can be calculated by characterizing each interval as either agreeing (both observers recorded behavior or no behavior) or disagreement (one observer recorded the behavior, while the other did not). The total number of agreements for the meeting is then divided by agreements plus disagreements and multiplied by 100 to obtain the average reliability for the entire observation. A fifth factor that could affect reliability and integrity deficiencies may be a risk of competing environmental uncertainties. Specifically, there may be reinforcements for protocol waivers, offenders for compliance with protocol or both. For example, a study by O`Leary, Kent and Kanowitz (1975) showed that observers who received specific information about the meeting (z.B.