The other variables in this series are the following:

• MTODMHI (male children 2-3) (available 1997 forward)

• FTODMHI (female children 2-3) (available 1997 forward)

• MKIDMHI (male children 4-11) (available 1997-2000)

• FKIDMHI (female children 4-11) (available 1997-2000)

• MTEENMHI (male children 12-17) (available 1997-2000)

Before asking the appropriate set of questions for female sample children age 12-17, the interviewer stated, "I am going to read a list of items that describe children. For each item, please tell me if it has been NOT TRUE, SOMETIMES TRUE, or OFTEN TRUE . . . [of sample child during the past 6 months]."

The Field Representative's Manuals for 1997-2000 instructed interviewers, "[i]f the respondent gives an answer which does not match the categories read in the question, reread the question emphasizing the wording of the answer categories."

The statements parents were asked to rate for female sample children age 12 to 17 included:

• Lies or cheats?

• Does poorly at school work?

• Has trouble sleeping?

• Has been unhappy, sad, or depressed?

Responses for these four items were coded by NHIS as 0 (Not True), 1 (Sometimes True), and 2 (Often True). Individual item scores were then summed to calculate an overall MHI score ranging from 0 (lowest need for mental health services) to 8 (highest need for mental health services), which is reported in FTEENMHI.

The set of statements that parents were asked to rate as the basis for MHI scores differed across child age groups, as well as differing between males and females in the same age group. For information on the items parents rated for other gender/age groups and for further details on the interpretation of MHI scores, see FTODMHI.

Beginning in 2001, the Strengths and Difficulties Questionnaire (SDQ) developed by Dr. Robert Goodman replaced the CBCL questions for children age 4-17. The extended SDQ fielded in 2001 and 2003-2004 included 25 questions divided among five subscales that measured emotional symptoms, conduct problems, hyperactive behavior, peer relationships, and prosocial behavior. In addition to the subscale items, eight questions were included as part of an impact supplement used to measure the duration and impact of the child's symptoms.

A shorter, 6-item version of the SDQ was fielded in 2002 and 2005 forward. The short SDQ included two items from the emotional symptoms subscale (WORRIED, UNHAPPY), one item from the conduct problems subscale (WELLBEHAVED), one item from the hyperactive behavior subscale (GOODATTEN), one item from the peer relationships subscale (GETALONGAD), and one item from the impact supplement (EMODIFF).

The public use files of the NHIS data include only the individual item scores for the SDQ subscale items. Analysts wishing to calculate the likelihood of a psychological problem for sample children age 4-17 using the SDQ variables may do so by summing responses for the appropriate subscale items. For more information on the SDQ, see any of the Child Strengths and Difficulties variables (e.g., BADTEMPER).