The results varied based on the experiment group type; for example in one study, the results reliability were between. 71 and . 86 and the study groups consisted of children in pediatric-medical outpatients, psychiatric referrals and public school students. Test – retest Reliability. The CDI is given to the individual and the questions are based on the feelings and function for the previous two weeks. The reason for using the prior two weeks is that the instrument is testing symptoms, not traits. The retest is within two to four weeks and has been found to have excellent short-term stability.
Standard Error. There are two types of standard errors related to CDI; “standard error of measurement (SEM1) and standard error of prediction (SEM2)” (Sitarenios & Kovacs, p. 272, 1999). The SEM2 is directly related to outcome assessment and is therefore an important element in retesting. This is one way in which to predict the level of change that would be found when retesting a respondent. Validity. The validity of the CDI demonstrates how reliable and accurate the results are and whether the results can be accurately estimated.
Validity is obtain when an instrument has been tested through various studies and can be verified through expected outcomes. The CDI has been tested in numerous studies, “and its validity has been well established using a variety of techiniques (Sitarenios & Kovacs, p. 273, 1999). The CDI has been found to be an appropriate instrument in detecting symptoms of depression in children ages seven to seventeen (1999). The main manual for the CDI is based on North American children and therefore should be taken into consideration when working with children outside of North America or children originally from outside the U.
S. (Sitarenios & Kovacs, 1999). There is specific research that has been conducted to assist clinicians in interpreting the CDI results for children from different ethnic background or countries. CDI Short Form The CDI short form is a ten question assessment used with children seven through seventeen and requires less time than the long form. This is a practical way of assessing the client when time is limited. When compared to the long form, research has shown the results are comparable (Sitarenios & Kovacs, 1999). Readability The readability of the CDI is often cited in the first grade level.
Sitarenios & Kovacs (1999) however tested the readability of the CDI among various grade levels using the Dale-Chall procedure in which the number of complete sentences [are] counted and divided into the number of words to determine the average sentence length (p. 275). Based on these calculations, a third grade reading level was determined to be the cited grade level for CDI; meaning the “often-cited first-grade reading level for the CDI is not definitive” (275). The implications of the readability test suggest that younger children may not understand the questions on the CDI.
Younger children, or those with reading problems should have the test read to them, while they read along, by the clinician to ensure comprehension. Administration Methods There are numerous ways in which the CDI can be administered. The following section discusses the ways in which a clinician can administer the test to children ages seven to seventeen. QuickScore. A QuickScore form may be used to administer the CDI to children. It is a self-contained form that allows the clinician to score and profile the CDI. The QuickScore is easily converted to T-scores and can be computer administered as well.
It may be decided that the question related to suicide is not appropriate for the individual or group, in which case the clinician may exclude the question or give the CDI short form which does not ask this question. The clinician must make sure that the child can comprehend the inventory. If there is any doubt, the clinician must read out loud the questions and responses. If the child does not feel any of the responses are a match for his feeling or experience than the clinician will ask him to choose the one that is the closest match. Group Administration. It is possible to give the test to a group rather than to an individual.
Applicable Populations Several variables must be considered when scoring the CDI. When interpreting the CDI, a clinician must consider the respondents “background, including their socioeconomic status, country of origin, and ethnicity” (Sitarenios & Kovacs, p. 276, 1999). Special consideration should be made when scoring the test of special circumstances such as “socioeconomic status, urban and rural residents, those in public housing situations, and children with mental retardation, learning disability, or emotional problems” (Sitarenios & Kovacs, p. 276, 1999).
Children of divorce, diabetic children and those with cancer or in remission should also get special consideration when interpreting CDI scores. Determining Results Validity When interpreting the CDI score, the clinician must seek out potential threats to the validity of the answers. There are two approaches to seeking validity of the CDI scores; determining the quality of the inventory and examining the Inconsistency Index. Quality of Inventory. The first step to determine the quality of the inventory is to make sure that all the questions have been filled out; any missing answers will affect the quality and accuracy of the inventory.
A value must be assigned to the missing answer(s) by averaging the remaining scores (Sitarenios & Kovacs, 1999). Next the clinician must determine whether there is bias present in the results. This can be achieved by examining the results; if all answers are the same, they may be bias, also if the results are conflicting this may be a sign of random answering and can be considered biased. The child may have unrealistic expectations of the CDI and feel that they must lie or deny certain feelings on the test. It is important than to discuss the expectations before giving the CDI.
The environment must be considered when determining the validity of CDI scores. Was the child in a safe, private environment, free of distractions, when taking the CDI? If not, the validity of the test may be compromised. Inconsistency Index. Rather than considering special items or scales, the inconsistency index allows the clinician to examine responses by determining if the responses are random or consistent. If there is a major discrepancy among correlating items, validation may be compromised. Nelson’s (et. al.
, 1987) study of the effectiveness of CDI focused on a less researched area; the effectiveness of CDI on emotionally disturbed children. The study focused on emotionally disturbed children, using variables such as sex, age and race to determine factor variations. The test subjects for Nelson (et. al. , 1987) consisted of 535 children newly admitted to one of two psychiatric hospitals; the Millcreek Psychiatric Center for Children in Cincinnati and the Virginia Treatment Center for Children in Richmond, VA. Within one week of their admission, children were given the CDI.
The majority of the test subjects were male and Caucasian. There were 305 adolescents and 220 preteens; ten subjects had missing data. The results of Nelson’s (et. el. , 1987) were less than favorable. Unfortunately the CDI was not the primary focus of the test; the test subjects were exposed to numerous other inventories, group meetings and experiences so that the actual validity of the CDIs were not reliable. Also, Nelson (et. al, 1987) chose to include eighteen year olds to the test subjects which is an older age than the CDI is intended for.