-------------------------------------------------------------------------------- -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- COR813 J.Yonker Graduate Mail Decision Rules Created: 01/06/2005 Updated: 04/22/2005 -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- Contents: A. General Mail Survey Comments B. Module Notes Index: 1. General 2. Health 3. Health Behaviors 4. Women's Health 5. World Health Organization (WHO) Vignettes 6. How You've Felt this Past Week 7. How You Feel During a Typical Week 8. Dealing with Problems 9. Values and Attitudes 10. Masculinity Items 11. Social Background 12. Marriage 13. Work and Family 14. Social Relationships 15. Social and Civic Participation 16. Religion and Spirituality C. Survey/Replicate Updates 1. Added Items 2. Changed Items 3. Corrected Items and Formatting Changes D. Data Use Suggestions E. Data Editing and Entry Protocol 1. Data Editor Specifications 2. Data Entry Specifications F. Additional References and Resources -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- A. General Mail Survey Comments: How to use this COR: Overview: Response Rate: (To be determined) - Although "Don't Know" was not an explicit response choice, we did allow write-in "Don't Know" if the respondent indicated such. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-1 General -------------------------------------------------------------------------------- Need to talk about replicat vs group91 disparity. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-2 Health -------------------------------------------------------------------------------- IXSF1REC, IXSF2REC SF-12 Physical and Mental Health Indices IXSF3REC Number of SF-12 items answered Component Variables: -GX201RE (x1): GH1* -IXCH01RE (q7h_a): PF02 -IXCH02RE (q7h_b): PF04 -IX401RER (q9h_a): RP2 -IX402RER (q9h_b): RP3 -IX403RER (q10h_a): RE2 -IX404RER (q10h_b): RE3 -IX405RER (q11h): BP2* -IX406RER (q12h_a): MH3* -IX407RER (q12h_b): VT2* -IX408RER (q12h_c): MH4 -IX409RER (q13h): SF2 *Items was reverse scored for IXSF1REC and IXSF2REC. All 12 items must have valid responses for IXSF1REC and IXSF2REC to be computed. Source: Ware, John E., Mark Kosinski, and Susan D. Heller. 1995. SF-12: How to Score the SF-12 Physical and Mental Health Summary Scales. Second Edition. Boston, MA: The Health Institute, New England Medical Center. -------------------------------------------------------------------------------- IX010REC R's total height in inches 1. There were two components to this item: A. IX010FRE: Feet B. IX010IRE: Inches 2. If the feet and inches components both contained valid responses, then feet were converted to inches and added to the inches component as one would expect. 3. If the feet component contained a valid response but the inches component was missing/blank, then the feet component was converted to inches and the inches component was assumed to be zero. 4. If the inches component contained a valid response but the feet component was missing or refused, no total height value was computed. 5. Associated variables: -IX010FRE: Height, feet -IX010IRE: Height, inches -IX011REC: Body Mass Index (BMI) -------------------------------------------------------------------------------- IXC16RER If yes (to items IXC12RER-IXC15RER), what was the most serious health condition that caused these problems? 1. Open ended responses were coded using the ICD-9-CM coding scheme. 2. Associated variables: -IXC01RER: R's childhood health rating. -IXC02RER-IXC12RER: R's childhood illnesses. -IXC13RER-IXC15RER: R's childhood health limitations. NOTE: If any item in (IXC13RER-IXC15RER) is 'Yes' there should be a valid code for IXC16RER. NOTE: Refer to ICD-9-CM coding scheme. -------------------------------------------------------------------------------- IXA01REC Does R know about the health of any biological parent or sibling? 1. Coding: 1 = Yes, R knew about at least one biological parent or sibling's health. 2 = No, R does not know about any biological parent or sibling's health. 2. Refusals and anything other than a 'Yes' (1) were counted as 'No' (2). 3. This variable should have no missing or negative values. 4. Associated variables: -IXF01RER-IXF12RER: Biological father's health. -IXM01RER-IXM12RER: Biological mother's health. -IXB01RER-IXB12RER: Biological brother's health, if any. -IXS01RER-IXS12RER: Biological sister's health, if any. -------------------------------------------------------------------------------- IXA02REC-IXA12REC Has any biological parent or sibling had [condition]? 1. These are dummy variables indicating whether any biological first-degree relative had one of the 11 selected illnesses/diseases. 2. As with IXA01REC, anything other than a 'Yes' (1) was scored as 'No' (2). 3. There should be no missing or negative values for these variables. 4. Associated variables: -IXA01REC: Does R know about any first-degree relative's health? -IXF02REC,IXM02REC,IXB02REC,IXS02REC: High blood pressure -IXF03REC,IXM03REC,IXB03REC,IXS03REC: High blood cholesterol -IXF04REC,IXM04REC,IXB04REC,IXS04REC: Stroke before age 65 -IXF05REC,IXM05REC,IXB05REC,IXS05REC: Stroke age 65 or older -IXF06REC,IXM06REC,IXB06REC,IXS06REC: Heart attack before age 55 -IXF07REC,IXM07REC,IXB07REC,IXS07REC: Heart attack age 55 or older -IXF08REC,IXM08REC,IXB08REC,IXS08REC: Diabetes -IXF09REC,IXM09REC,IXB09REC,IXS09REC: Alzheimer's -IXF10REC,IXM10REC,IXB10REC,IXS10REC: Asthma -IXF11REC,IXM11REC,IXB11REC,IXS11REC: Osteoporosis -IXF12REC,IXM12REC,IXB12REC,IXS12REC: Cancer -IXF13REC,IXM13REC,IXB13REC,IXS13REC: Cancer (Specify Type 1) -IXF14REC,IXM14REC,IXB14REC,IXS14REC: Cancer (Specify Type 2) -IXF15REC,IXM15REC,IXB15REC,IXS15REC: Cancer (Specify Type 3) -------------------------------------------------------------------------------- IXF13RER,IXM13RER,IXB13RER,IXS13RER IXF14RER,IXM14RER,IXB14RER,IXS14RER Please use the lines in each column to indicate the name of the organ or system where the cancer occurred. 1. The cancerous organs/systems indicated by the respondent were recoded into the ICD-9-CM coding scheme. 2. Up to two organs/systems were coded per relative, one for each blank provided per on the questionnaire. 3. If the respondent indicated more than two organs/systems for a given relative,[IXF15RER,IXM15RER,IXB15RER,IXS15RER] was set to 'Yes' (1), respectively, to indicate that additional organs/systems were mentioned for that relative. 4. Associated variables: -IXF12RER,IXF15RER: Father (1) ever, (2) more than two types indicated -IXM12RER,IXM15RER: Mother (1) ever, (2) more than two types indicated -IXB12RER,IXB15RER: Brother (1) ever, (2) more than two types indicated -IXS12RER,IXS15RER: Sister (1) ever, (2) more than two types indicated -------------------------------------------------------------------------------- IXSP01RE If you are married, would you say that your spouse's health is excellent, very good, good, fair or poor? 1. The response choice of 'Not Married' (6) was collapsed into 'Inap, Not Married' (-2) for ease of use. There was the possibility that users would not recognize this category and inlcude it in analyses. 2. The item was reverse coded to match the convention of greater health associated with higher scores. 5 = Excellent 4 = Very Good 3 = Good 2 = Fair 1 = Poor 3. If the respondent indicated that they were NOT married in the marriage module, any response to this item was overridden and became 'Inap' (-2). 3. Associated variables: -MX001RER,GX001RER: R's current health rating. -MX002RER,GX002RER: R's health rating compared with others same age/sex. -MX003RER,GX003RER: R's health compared to 10 years ago. -MX004RER,GX004RER: R's appearance compared to 10 years ago. NOTE: Some associated variables have alternate response categories. NOTE: Item was added starting in replicate 5 (GROUP91>4). -------------------------------------------------------------------------------- IX016RER-IX027RER IX031RER-IX039RER IX046RER-IX081RER IX121RER-IX141RER 1. There were two components to each physical symptom: A. How often have you had [symptom] in the past six months? B. How much discomfort has [symptom] caused you in the past six months? 2. Variable numbering is non-sequential because some items were also asked in 1992 and thus retained their original variable names. A. Items asked in 1992 and 2003 (18): -Aching muscles -Back pain/strain -Chest pain -Constipation -Diarrhea -Dizziness/faintness -Excessive sweating -Fatigue/exhaustion -Headache -Lack of energy -Numbness -Ringing in ears -Shortness of breath -Skin problems -Stiff/swollen joints -Trouble sleeping -Upset stomach -Urination problems B. Items added in 2003 (7): -Bone pains -Coughing/wheezing -Difficulties with or painful sexual intercourse -Neck and/or shoulder pain -Pain in hands/wrists -Pain in ankles/knees -Palpitations 3. If the respondent indicated in the first component that they had not experienced [symptom] in the past six months and the second component was left blank, then the second component was set to 'Inap' (-2). 4. If the respondent indicated in the first component that they had not experienced [symptom] in the past six months and indicated in the second component that [symptom] caused no discomfort, then the second component was set to 'Inap' (-2). 5. Associated variables: -IX082REC: Total number of physical symptoms. -IX082RED: Total number of physical symptom questions answered. -IX117REC: Total number of diagnosed illnesses. -IX117RED: Total number of illness qeustions answered. -------------------------------------------------------------------------------- IX085RER-IX086RER IX089RER-IX090RER IX097RER-IX098RER IX103RER-IX112RER IX142RER-IX153RER 1. There were three components to each of the 25 selected illnesses: A. Has a medical professional ever diagnosed you with [illness]? B. How old were you when you were first diagnosed with [illness]? C. How much does [illness] currently interfere with what you like to do? 2. Variable numbering is non-sequential because some items were also asked in 1992 and thus retained their original variable names. A. Items asked in 1992 and 2003 (6): -Asthma -Chronic bronchitis/Emphysema -Circulation problems -Kidney/bladder problems -Serious back trouble -Ulcer B. New items added in 2003 (8): -Allergies -Chronic sinus problems -Fibromyalgia -High cholesterol -Irritable bowel syndrome -Multiple sclerosis -Osteoporosis -Prostate problems (men) 3. To retain the numbering from 1992 when only two questions were asked per illness, we created -RER and -ARE suffixes. The -ARE variables indicate the age of the respondent when they were first diagnosed with [illness]. The -RER variables are analagous to the 1992 versions. 4. If the respondent indicated an age of initial diagnosis or indicated at least some interference in the ability to do what he/she likes, the 'Ever diagnosed' component (A) was set to 'Yes' (1) regardless of what the respondent indicated. 5. If the respondent indicated that they had never been diagnosed with [illness] and left both the age and interference components blank (B and C, respectively), then both of these components were reset to 'Inap' (-2). 6. If the respondent indicated that they had never been diagnosed with [illness], left the age blank, and answered 'None' (1) for the interference component, then both the age and interference components (B and C) were reset to 'Inap' (-2). 7. The 'Prostate Problems' item was automatically set to 'Inap' for female respondents. 8. Associated variables: -IX117REC: Total number of diagnosed illnesses. -IX117RED: Total number of illness questions answered. -IX082REC: Total number of physical symptoms. -IX082RED: Total number of physical symptom questions answered. -------------------------------------------------------------------------------- IXHS02RE-IXHS05RE IXHS06RE-IXHS07RE What type of vision correction do you regularly use? 1. There are four components to this question: A. Prescription glasses B. Prescription contact lenses C. Non-Prescription magnifying glasses D. Any vision corrections 2. The double negative wording of the 'No vision corrections' component was reversed to indicate whether the respondent uses 'Any' vision corrections. 1 = Yes, R uses vision corrections 2 = No, R does NOT use vision corrections 3. If the respondent indicated any type of vision correction (A-C) but also indicated 'No' (2) to 'Any vision corrections' then the latter was reset to 'Yes' (1) and the dummy variable IXHS07RE was set to 'Yes' to indicate a problematic response pattern. 4. A single variable (IXHS06RE) was constructed to summarize answers to the series. -3 = Refused all items in series -2 = Inap, GROUP91 < 5 1 = Rx Glasses 2 = Rx Contact Lenses 3 = Rx Glasses and Contact Lenses 4 = Non-Rx Magnifying Glasses 5 = Rx Glasses and Non-Rx Magnifying Glasses 6 = Rx Contact Lenses and Non-Rx Magnifying Glasses 7 = Rx Glasses and Contact Lenses and Non-Rx Magnifying Glasses 8 = No vision corrections 5. Associated variables: -IXHS01RE: How often do you have your eyes examined? -IXHS08RE-IXHS20RE: Hearing and Hearing Problems -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-3 Health Behaviors -------------------------------------------------------------------------------- IDIFFREC An indicator variable for whether R answered 'Yes' to any item in the 'Difficulty Obtaining Healthcare' series. (q38hb_:a-k) There should be no missing values for this item. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-6 How You've Felt this Past Week -------------------------------------------------------------------------------- IU001REC,IU002RE Modified CES-D Summary, # of Questions Answered IU003RER-IU022RER Modified CES-D items IU023RER-IU025RER Hostility Index (NSFH2) IU036RER-IU048RER Modified Speilberger Anger & Anxiety Index (MEMO144) (NSFH2) IUA33REC,IUB33RE Summary Index for Speilberger Anxiety Index IUA34REC,IUB33RE Summery Score for Speilberger Anger Index -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-8 Dealing with Problems -------------------------------------------------------------------------------- ID001ARE-ID017ARE: First or only time [event] happened: Age? ID001BRE-ID017BRE: Last time [event] happened: Age? ID001CRE-ID017CRE: Has [event] ever happened? 1. All ages were rounded down. A responded that was 64 and 1/2 years old when [event] happened was considered to be simply 64 years old. Occassionally, a respondent indicated that [event] occurred when they were less than 1 year old. To avoid confusion, we parted from the general rule and these were rounded up to 1 year. 2. If a valid answer was given to either the first/only or last time [event] happened, including 'Dont Know' (-1), the 'Ever' component was set to 'Yes' (1) regardless of what the respondent indicated. 3. If a valid answer was given to the first/only time [event] happened, including 'Dont Know' (-1), and the last time [event] happened was blank, the last time was reset to 'Inap' (-2) and it was assumed that [event] only happened once. 4. If the respondent indicated that [event] never happened and the first/only and last compents were left blank, the first/only and last time components were set to 'Inap' (-2). 5. If the respondent indicated a valid age for the last time [event] happened, including 'Dont Know' (-1), but the first/only time [event] happened was blank, the first/only time component was set to 'Refused' (-3). 6. To avoid user confusion, the double negative wording of the 'Never' raw variable and response category combination was reversed to indicate whether the respondent 'Ever' experienced [event]. 7. Some respondents indicated a younger age the last time [event] occurred than the first/only time. These were not corrected, but the variable [IX001DRE- IX017DRE] was set to 1 to indicate a problem with the age pattern of that particular event. 1 = Respondent indicated an illogical age pattern 2 = Respondent indicated (1) logical age pattern, (2) event did not occur, or (3) event did not occur more than once. 8. The variable nomenclature followed the order that they appeared in the mail survey. For example, the 'Ever' component came last in the triplet of questions pertaining to each event and thus has the suffix of 'C'. 9. Associated variables: -ID001DRE-ID017DRE: Respondent indicated illogical age pattern for [event]. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-11 Social Background -------------------------------------------------------------------------------- IE008RE-IE018RE What is your race or origin? 1. There were six distinct races/origins and a fill-in 'Other' category the respondent could choose: A. White B. Black, African-American, or Negro C. Asian D. Indian (American) or Alaskan Native E. Pacific Islander F. Hispanic, Latino, or Spanish Origin G. Some other race 2. Respondents were free to select any group or combination of groups. 3. If a respondent indicated that they belonged to at least one race/origin (1) then the remaining blank check boxes were assumed to be 'No' (2). 4. If the respondent left all of the boxes blank, then the entire series was considered a refusal and all items were set to 'Refused' (-3). 5. Associated variables: -IE019REC: Total number of races/origins respondent indicated. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- B-12 Marriage -------------------------------------------------------------------------------- IC008RER,IC013RER Reference: Haynes, S.N., Floyd, F.J., Lemsky, C., Doll, E.R., Winemiller, D., Heilman, N., Werle, M., Burgy, T.M., & Cardone, L. (1992). The Marital Satisfaction Questionnaire for Older Persons. Psychological Assessment, 4, 473-482. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- C-1 Item Additions -------------------------------------------------------------------------------- NOTE: All values shown reflect raw variable response choices and are NOT necessarily included in final created variables. q18wh (IX154RER): -Have you signed an organ donor card or indicated on your driver's license you intend to be an organ donor? 1 = Yes 2 = No -Added beginning with replicate four (GROUP91>3). -Added to 'Health' module. q14a (IXSP01RER): -If you are married, would you say that your spouse's health is excellent, very good, good, fair, or poor? 1 = Excellent 2 = Very Good 3 = Good 4 = Fair 5 = Poor 6 = Not Married -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q24 (IXHS01RE): -How often do you have your eyes examined? 1 = Every year or more often 2 = Every 1 to 2 years 3 = Less often than every 2 years 4 = Never -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q25 (IXHS02RE-IXHS07RE): -Which type of vision correction do you regularly use? (Check ALL that apply.) 1 = Prescription glasses 2 = Prescription contact lenses 3 = Non-Prescription magnifying glasses 4 = No vision corrections -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q26 (IXHS08RE): -Have you had your hearing checked in the past 5 years? 1 = Yes 2 = No -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q27 (IXHS09RE): -Do people that live with you or are close to you ask you whether you think that you should have your hearing checked? 1 = Yes 2 = No -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q28 (IXHS10RE): -Which of the following best describes your use of hearing aids? 1 = I have hearing aids for one or both ears and use them regularly 2 = I have hearing aids for one or both ears but do not use them regularly 3 = I do not own hearing aids -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q29_:a-e, q29b_:b-e (IXHS11RE-IXHS20RE): -Individual Items: A. Hearing conversations in person B. Hearing conversations on the phone C. Understanding spoken instructions from your doctor, employer, or other person D. Reading small print on medicine bottles or other places E. Understanding written instructions -There were two components to each item: -(Have these) Problems increased in the last 12 months? 1 = Yes 2 = No -(Have these) Problems increased in the last 5 years? 1 = Yes 2 = No -Added beginning with replicate five (GROUP91>4). -Added to 'Health' module. q1a_:a-k (IM01RER-IM10RER): -Individual Items: A. It is important for a man to have a male friend he can confide in. B. When a husband and wife make decisions about buying major things for the home, the husband should always have the final say. C. A man should always project an air of confidence even if he really doesn't feel confident inside. D. A working mother can establish just as warm and secure a relationship with her children as a mother who does not work. E. It bothers me when a man does something that I consider 'feminine.' F. A husband whose wife is working full-time should spend just as many hours doing housework as his wife. G. Men have greater sexual needs than women. H. When a man is feeling pain he should not let it show. I. In some kinds of situations a man should be ready to use his fists. J. It is important for a woman to have a female friend she can confide in. K. Being larger, stronger-looking, and more muscular makes men more attractive to women. -Response Choices: 1 = Strongly Agree 2 = Agree 3 = Neither Agree nor Disagree 4 = Disagree 5 = Strongly Disagree -Added beginning with replicate five (GROUP91>4). -Added to the 'Values and Attitudes' module. q5vaff (IN031RER): -I enjoy personal and mutual conversations with family members and friends. 1 = Agree Strongly 2 = Agree Moderately 3 = Agree Slightly 4 = Disagree Slightly 5 = Disagree Moderately 6 = Disagree Strongly -Added beginning with replicate five (GROUP91>4). -Added to the 'Values and Attitudes' module. abuse (Not for public release, no public variable created): -If you would like to clarify or tell us more about any emotional, physical, or sexual abuse you experienced as a child, please use this space. -Added beginning with replicate four (GROUP91>3). -Added to the 'Social Relationships' module. Q6sp:d_1-h_3 (IZ117RER-IZ131RER) -There were three components to each item. -During the past year, I did I did this: [blank] hours per week. -How often did you do this about 10 years ago? 1 = Often 2 = Rarely 3 = Never -How often did you do this when you were about 35? 1 = Often 2 = Rarely 3 = Never -Individual Items: -Reading on the job -Reading biographies or other non-fiction books -Reading the Bible or other religious materials -Reading magazines or newspapers -Reading fiction -Added beginning with replicate five (GROUP91>4). -Added to the 'Social and Civic Participation' module. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- C-2 Item Changes -------------------------------------------------------------------------------- World Health Organization Vignette Changes: NOTE: See also MEMO145. Variables: -Raw: -q20_:a-b -q21_:a-b -q22_:a-b -q23_:a-b -Constructed: -IX511RER-IX513RER -IX521RER-IX523RER -IX531RER-IX533RER -IX531RER-IX533RER -IX531RER-IX533RER -IX531RER-IX533RER -IX531RER-IX533RER -IX531RER-IX533RER Mobility Vignettes (4): Vignette A: [Name] is able to walk distances of up to 1/8 mile without any problems but feels tired after walking 1/2 mile or climbing up more than one flight of stairs. He [She] has no problems with day-to-day physical activities, such as carrying food from the market. Vignette B: [Name] does not exercise. He [She] cannot climb stairs or do other physical activities because he [she] is obese. He [She] is able to carry the groceries and do some light household work. Vignette C: [Name] has a lot of swelling in his [her] legs due to his health condition. He [She] has to make an effort to walk around his home as his [her] legs feel heavy. Vignette D: [Name] is able to move his [her] arms and legs, but requires assistance in standing up from a chair or walking around the house. Any bending is painful and lifting is impossible. Mobility Questions (2): 1. Overall, how much of a problem did [name] have with moving around? 2. How much difficulty did [name of person] have in vigorous activities, such as running 2 miles or cycling? Affect Vignettes (4): Vignette A: [Name] enjoys his [her] work and social activities and is generally satisfied with his [her] life. He [she] gets depressed every 3 weeks for a day or two and loses interest in what he [she] usually enjoys but is able to carry on with his [her] day to day activities. Vignette B: [Name] worries often about his [her] health. He [She] gets depressed once a week for a day or two, thinking about what could go wrong and all the illnesses he [she] could get, but is able to come out of this mood if he [she] concentrates on something else. Vignette C: [Name] feels nervous and anxious. He [She] worries and thinks negatively about the future, but feels better in the company of people or when doing something that really interests him [her]. When he [she] is alone he [she] tends to feel useless and empty. Vignette D: [Name] feels depressed most of the time. He [she] weeps frequently and feels hopeless about the future. He [She] feels that he [she] has become a burden on others and that he [she] would be better dead. Affect Questions (2): 1. How much of a problem did [name] have with feeling sad, low, or depressed? 2. How much of a problem did [name] have with worry or anxiety? Response Categories (All Questions): 1 = None 2 = Mild 3 = Moderate 4 = Severe 5 = Extreme Male Names (Only given to male respondents, GENFORM = 1): 1. Robert 2. James 3. Richard 4. John Female Names (Only given to female respondents, GENFORM = 2): 1. Mary 2. Judith 3. Carol 4. Barbara Replicates 0,4,8 (GROUP91 = 0,4,8): 1. IX511RER-IX513RER -Affect Domain -Vignette B -Name 4 2. IX521RER-IX523RER -Mobility Domain -Vignette A -Name 3 3. IX531RER-IX533RER -Affect Domain -Vignette C -Name 2 4. IX541RER-IX543RER -Mobility Domain -Vignette D -Name 1 Replicates 1,5 (GROUP91 = 1,5): 1. IX551RER-IX553RER -Mobility Domain -Vignette C -Name 1 2. IX561RER-IX563RER -Affect Domain -Vignette D -Name 2 3. IX521RER-IX523RER -Mobility Domain -Vignette A -Name 3 4. IX511RER-IX513RER -Affect Domain -Vignette B -Name 4 Replicates 3,6,9 (GROUP91 = 3,6,9): 1. IX541RER-IX543RER -Affect Domain -Vignette A -Name 2 2. IX571RER-IX573RER -Mobility Domain -Vignette B -Name 1 3. IX561RER-IX563RER -Affect Domain -Vignette D -Name 4 4. IX551RER-IX553RER -Mobility Domain -Vignette C -Name 3 Replicates 2,7 (GROUP91 = 2,7): 1. IX511RER-IX513RER -Mobility Domain -Vignette D -Name 3 2. IX531RER-IX533RER -Affect Domain -Vignette C -Name 4 3. IX571RER-IX573RER -Mobility Domain -Vignette B -Name 1 4. IX541RER-IX543RER -Affect Domain -Vignette A -Name 2 -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- C-3 Item Corrections and Formatting Changes -------------------------------------------------------------------------------- Formatting Changes: Male Form: 1. Pages 8,9,44,45,46,47 (Now Pages 8,9,45,46,47,48) -Replace vertical daches between columns -Changed beginning with replicate 5 (GROUP91>4) 2. Page 10 (Now Page 11) -Reformat race items to make room for new items on page -Move 'Values and Attitudes' module heading to page 11 -Changed beginning with replicate 5 (GROUP91>4) 3. Pages 11-50 (Now Pages 12-51) -Change page numbers -Changed beginning with replicate 5 (GROUP91>4) 4. Pages 16,18,19,20 (Now Pages 17,19,20,21) -Change skip instructions (No: Please go to Question 1 on Page 22) -Changed beginning with replicate 5 (GROUP91>4) 5. Page 17 (Now Page 18) -Capitalize and bold 'FIRST' employer retired from and reformat to allow more space for employer name -Changed beginning with replicate 4 (GROUP91>3) 6. Page 32 (Now Page 33) -Change skip instructions (No: Please go to Qeustion 9 on Page 34) -Changed beginning with replicate 5 (GROUP91>4) 7. Page 41 (Now Page 42) -Change skip instructions (No: Please go to Qeustion 6 on Page 43) -Changed beginning with replicate 5 (GROUP91>4) 8. Page 42 (Now Page 43) -Change skip instructions (No: Please go to Qeustion 1 on Page 44) -Changed beginning with replicate 5 (GROUP91>4) 9. Page 45 (Now Page 46) -Reformat page to allow for new items and moving Q7d to next page -Changed beginning with replicate 5 (GROUP91>4) 10. Page 46 (Now Page 47) -Reformat with addition of item Q7d -Changed beginning with replicate 5 (GROUP91>4) 11. Page 51 (Now Page 52) -Keep e-mail question without Q19 label -Changed beginning with replicate 5 (GROUP91>4) Female Form: 1. Page 10 -Change note at top of page: Questions 15-17 concern men's health... -Changed beginning with replicate 5 (GROUP91>4) 2. Pages 10,11,46,47,48,49 (Now Pages 10,11,47,48,49,50) -Replace vertical daches between columns -Changed beginning with replicate 5 (GROUP91>4) 3. Pages 12-52 (Now Pages 13-53) -Change page numbers -Changed beginning with replicate 5 (GROUP91>4) 4. Page 12 (Now Page 13) -Reformat race items to make room for new items on page -Move 'Values and Attitudes' module heading to page 13 -Changed beginning with replicate 5 (GROUP91>4) 5. Pages 18,20,21,22 (Now Pages 19,21,22,23) -Change skip instructions (No: Please go to Question 1 on Page 24) -Changed beginning with replicate 5 (GROUP91>4) 6. Page 19 (Now Page 20) -Capitalize and bold 'FIRST' employer retired from and reformat to allow more space for employer name -Changed beginning with replicate 4 (GROUP91>3) 7. Page 34 (Now Page 35) -Reformat to allow for new item -Changed beginning with replicate 4 (GROUP91>3) 8. Page 34 (Now Page 35) -Change skip instructions (No: Please go to Question 1 on Page 36) -Changed beginning with replicate 5 (GROUP91>4) 9. Page 37 (Now Page 38) -Change skip instructions (No: Please go to Question 1 on Page 39) -Changed beginning with replicate 5 (GROUP91>4) 10. Page 43 (Now Page 44) -Change skip instructions (No: Please go to Question 1 on Page 45) -Changed beginning with replicate 5 (GROUP91>4) 11. Page 44 (Now Page 45) -Change skip instructions (No: Please go to Question 1 on Page 46) -Changed beginning with replicate 5 (GROUP91>4) 12. Page 47 (Now Page 48) -Reformat page to allow for new items and moving Q7d to next page -Changed beginning with replicate 5 (GROUP91>4) 13. Page 48 (Now Page 49) -Reformat with addition of item Q7d -Changed beginning with replicate 5 (GROUP91>4) 14. Page 54 -Keep e-mail question without Q19 label -Changed beginning with replicate 5 (GROUP91>4) Dropped Items: Male Form: -Page 51 -Drop entire page (paragraphs A-C) -Drop Q17: Who are you most like? -Drop Q18: Who are you least like? -Changed beginning with replicate 5 (GROUP91>4) Female Form: -Page 53 -Drop entire page (paragraphs A-C) -Drop Q17: Who are you most like? -Drop Q18: Who are you least like? -Changed beginning with replicate 5 (GROUP91>4) -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- D. Data Use Suggestions -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- E. Data Editing and Entry Protocol -------------------------------------------------------------------------------- E-1: Data Editor Specifications P9353: WLS Mail Survey The editor shall be the first to receive the returned, completed surveys once they have been removed from their business-reply envelopes and updated as returned in the master database. The editor will then, with a sharp eye, use these specifications to mark up the hard-copy surveys. The editor should use a RED pen or marker for ALL editing on the mail survey. This will allow anyone to distinguish between the respondent's answers and those added, deleted, or changed by the editor. Be sure the Respondent followed the correct skip pattern by answering the appropriate questions. Then make marks using the following specifications. If you're unclear about the way the Respondent filled out this section, mark the page with a post it note and set the questionnaire aside to be reviewed by the project director. 1. If the respondent uses some other symbol or system for marking: If the respondent draws lines through answer categories, or using some other system of marking their answers, use your judgment to decide whether the respondent is trying to indicate that this is how they want to answer or if they are trying to indicate that they do not like the available answer categories. The respondent may just be idiosyncratic or they may be indicating some level of refusal. If you think the respondent may be indicating a refusal OR if you are unsure either way, then mark the page with a post it note and set the questionnaire aside for review by the project director. If you think that the respondent is simply using some other marking to actually answer the question (is not indicating a refusal), then use your red pen to properly fill in the answer. For instance, check the box 'Yes' if the respondent has done something strange like draw a line through the 'Yes' box. 2. If the respondent leaves extraneous comments or uses own marking system: Draw an arrow to point to the comment or marking and then write 'F1 Note'. This will alert the data entry person to leave a note about this. If it is an extraneous comment, then the data entry person will simply end the comments verbatim. But if it is a matter of the respondent using their own marking system for answering the question(s), then write F1 Note 'Respondent used own marking system.' 3. An individual question or part of a question is left blank: Write the letter "b" next to the question number/letter in red. The data entry person will then be alerted to code the missing response as "Blank" 4. An entire section or page is left blank that SHOULD HAVE BEEN filled out Write the letter "b" next to the question number/letter in red. Draw a line from the 'b' down the one side of the page to indicate that the whole page is blank. The data entry person will then be alerted to code the missing responses (however many items are on the page) as "Blank" 5. Multiple answers are given to questions that require only 1 answer and only 1 answer can be entered Flag the question (or part of the question) by underlining the question number/ letter. The data entry person will then be alerted to follow the instructions on entering multiple responses. 6. For the first 'Dealing with Problems' question on page 23, question 1 for females and page 21, question 1 for males. If the respondent gives an age for either or both of the columns, AND checks Never, then cross out the Never answer and circle the age or ages. Leave an F1 Note alert 'R answered age and Never; age entered as answer' 7. If answers are given to questions that were supposed to be skipped If the respondent should have skipped an item or a series of questions given their answer to an earlier question, then you will actually change their answer to that earlier question to make it consistent with the following questions. For instance, if the respondent is asked 'Have you ever smoked cigarettes regularly in your entire life?' and the respondent answers 'No' (and is thereby supposed to skip a whole series of questions), but then proceeds to answer all of the following questions about smoking, then you will actually change their original 'No' answer to 'Yes'. The most important thing to keep in mind is that a judgment may need to be made in doing this so that the least amount of data is sacrificed. The purpose of this spec is to include (data enter) as much of the respondent's answers as possible. If following this spec causes more data to be lost than gained, then the spirit of the spec is not being followed. A judgment call will need to be made to allow data entry staff to follow a path in the instrument that includes the most data. The editor should make this judgment or defer to the project director if unsure. In the latter situation, place a post-it on the page and file the survey with problem cases. After making this answer change, leave the mark 'F1 note' to alert the data entry person to leave a note about this. The editor should think about the item where the data entry will person will have to leave the note. For instance, the note may be in regards to an item that the instrument will automatically skip over. Therefore, the editor should decide what is the next best item to leave the note at and indicate this with the red pen. Usually the best item to leave the F1 note at will be on the same page as the item about which the note is being left. Specifically, for Q6 in the Marriage section of the Male version 'Do you have a sexual partner?' - if this item is answered when it should have been skipped, always leave the F1 Note alert at Q7. This prevents the DE person from having to flip pages to understand the situation. 8. The answer is marked in between two questions or between answer boxes Underline the respondent's answer/marking as an alert for DE staff to use the spec for multiple answers. 9. Inches with decimals or fractions Round it to the nearest quarter inch (e.g. 2.25). If it is already reported like that, just leave it. If the respondent writes 2 and 1/4 inches then you will need to convert it to a decimal = 2.25 inches. 10. Age with month or fractions The age should be entered in whole numbers only, so cross out any months or fractions of a year written in for the age. (Their age should be the number of years old they were on their last birthday.) 11. A range is given instead of one number Take a midpoint. If the midpoint is NOT a whole number, round UP on odd numbers and DOWN on even numbers. For example, with a range of 15-20 years the midpoint is 17.5 so the answer would be 18 years. For a range of 10-15 years, the midpoint is 12.5 so the answer would be 12 years. In order to calculate the midpoint, add together the two numbers of the range and then divide by two. (This is similar to finding the average of two numbers.) For instance, add 15 + 20 = 35 / 2 = 17.5. For the midpoint, since 17.5 is an odd number, round up to 18. Write next to the question: 'Midpoint: 12' Next, simply re-write the range into the data entry format: x/x//, where x represents the low end and high end of the range. For instance if the respondent answers 15-20 then use your red pen to re- write it as 15/20//. This alerts the data entry staff to enter it this way. 12. For the following series of questions (Page 7 if male and page 6 if female), 16. Has a medical professional ever said you have any of the illnesses or conditions listed below? 16a. How old were you when first diagnosed with this illness or condition? 16b. How much does each of your illnesses or conditions currently interfere with what you like to do? The respondent answered 16 but not 16a/16b or answered 16a/16b but not 16: Treat it as blank and write the letter "b" next to the question 13. Kind of help GIVEN, kind of help RECEIVED, and Kind of help you could ask for (Questions 4-5, page 29-30 if male, page 31-32 if female) This also applies to obtaining Internet question on page 51, question 10 in female and page 49, question 10 if male. If the categories "None of these people needed help," "Help not needed," or "No one" or 'Not True' is checked, nothing else for that item (i.e. in that row) should be checked. Anything else that is checked should be crossed out with red pen. Essentially, the editor needs to make the respondent's answers consistent here. If the respondent writes words in the boxes referring to their spouse (husband or wife), or young children, cross it out. These questions EXCLUDE spouses and young children. Then write ' F1 note' next to these boxes alerting the data entry person to enter an F1 note. 14. My immediate boss or supervisor / Other people at work (Questions 7 and 10, page 20 if female or page 18 if male) In the past, when the respondent circled 5 for the first item, the instrument would not allow an answer of 5 on the next three items. Please note that the instrument has been changed and now allows an answer of 5 for all of these items. The instrument is now properly behaving here. 15. Out-of-pocket spending on medical care (Question 37, page 40 if female or page 38 if male) If the respondent draws a line or a dash on the answer line, check the rest of the booklet to see if the R has left similar lines or dashes that clearly indicate zero or none. If this clear, then code this question answer similarly as a zero. If, when checking the rest of the booklet, no similar markings are found, then leave a 'b' for blank and an F1 Note alert 'R wrote in a dash' 16. When a respondent has scratched out an answer If the respondent answered one way, then scratched it out and answered another way, or if they scratched out their answer and did not fill out another one, write 'scratch' next to the item. If the respondent did not fill out another answer, then the editor will also need to leave a 'b' for blank. This alerts the data entry person to follow the spec for entering F1 notes about a scratched-out answer. 17. Any other problems with the way a survey is filled out Write your comments/questions on a post it note with the question number and page, and set it aside for review by the project director. 18. R doesn't give definite answer When the R answers something like 'approx 4' and does not give a specific answer, use your red pen to mark '4' for the de person and write 'F1 NOTE'. Editors must make their own notation on the bottom right back of the hardcopy survey indicating 'Ed' for editor then initials and date in legible handwriting. Revised 3/5/04 K.D. -------------------------------------------------------------------------------- E-2: Data Entry Specifications P9353: WLS Mail Survey 1. Note Function Key ( , the letter "n", then ) Data entry staff should leave a note using the note function key regarding every marking left by the editor for every survey. The only exception to this rule is when the editor's note is , in which case just code accordingly. All of the comments made by the respondent should be entered verbatim using the note function key. If the respondent comments include expletives (profanity) or some words in all capital letters for emphasis, then you may type this exactly. If the respondent's handwriting simply looks like it is all-caps, but not for emphasis, then do not enter the comment in the notes field as all caps. In some instances you will need to enter two separate notes regarding one particular item. For example, if the respondent answers in a range, you will need to enter one note that is the numbers of the range 15/20//, then a second note that is standard: 'Range//'. See below for further details. 2. Consistent Typing The client emphasizes that we be especially consistent in our entry of the other-specify and verbatim answers. For example, a respondent was born in Madison, not madison or MADISON. Pay attention as well when entering age, inches, and ranges (see below). 3. Typing Open-ended Answers Please be aware that when you need to leave an F1 Note regarding an item that has an open-ended response field, you should type the open-ended response first, then go into the F1 Note. Data entry staff have found that if you try to type the F1 Note first, it places that text into the actual open-ended answer field. For obvious reasons, this must be avoided. 4. "Don't Know" or "Refused" The only time the answer is coded "Don't Know" or "Refused" is when the respondent has left some written comment to this effect. If there is no such marking from the respondent for "Don't Know" or "Refused," then code it as "Blank." 5. Questions are left blank (not answered) A. If it is an individual question or part of a question: Enter the code for " Blank" and move on to the next question. Question numbers/letters with missing responses should have the letter "b" written next to them (Editor's marking). B If an entire section or page is left blank that should have been filled out: Enter at the primary question at the beginning of the section. This may automatically skip you through the next items. If not, you will need to enter a for each item to indicate that the entire question/page/section is not filled out. 6. If answers are given to questions that were supposed to be skipped If the respondent should have skipped an item or a series of questions given their answer to an earlier question, then the editor should have changed their answer to that earlier question to make it consistent with the following questions. For instance, if the respondent is asked 'Have you ever smoked cigarettes regularly in your entire life?' and the respondent answers 'No' (and is thereby supposed to skip a whole series of questions), but the proceeds to answer all of the following questions about smoking, then the editor should have changed their original 'No' answer to 'Yes'. You will simply enter according to the editor's marking and the respondent's following answers. Then you should leave an F1 note to explain this. This note should be standard-always type 'Editor answer change//' 7. Multiple answers are given to questions that require only 1 answer and only 1 answer can be entered: A. Note Function Key: Press to get into command mode, then hit the letter "n" and hit B. Type in the multiple responses. Separate each response by a "/" and followed by 2 forward slashes (e.g. 2/3/4// ) C. Hit. Then you will need to leave a second note. D. Note Function Key: Press to get into command mode, then hit the letter "n" and hit E. This note should be standard - always type 'Multiple answer//' F. Hit G. Enter the LOWEST response (as in the lowest code) as the answer to the question. For example, take 2 if the respondent answered 2 and 3 and 4. H. Hit and move on to the next question NOTE: Any questions that have multiple responses should have been underlined in RED during the editing phase. 8. The answer is marked in between two questions If a check or a circle, or any marking, is placed between two questions, or between answer boxes, you will use the same spec as for entering multiple answers. A. Note Function Key: Press to get into command mode, then hit the letter "n" and hit B. Type in the answer categories between which the respondent's marking was made in the note function: Separate each response by a "/" and followed by 2 forward slashes (e.g. very good/excellent//) C. Hit. Then you will need to leave a second note. D. Note Function Key: Press to get into command mode, then hit the letter "n" and hit E. This note should be standard--always type 'Between answer//' F. Hit G. Enter the LOWEST response (as in the lowest code) as the answer to the question. For example, take 'very good' if the respondent answered between 'very good' and 'excellent.' H. Hit and move on to the next question 9. Need to backup to the previous screen to re-enter the correct answer A. Hit B. Leave an F1 Note. Type "ca [correct answer in this situation]" (e.g. ca 2 ) in the F1 Note C. Enter the correct answer (e.g. 2) D. Hit to go back to the question you were on before you backed up. 10. Inches with decimals or fractions Round it to the nearest quarter inch (e.g. 2.25). If it is already reported like that, just leave it. If the respondent writes 2 and 1/4 inches then you will need to convert it to a decimal = 2.25 inches. 11. Age with month or fractions The age should be entered in whole numbers only, so the Editor should have crossed out any months or fractions of a year written in for the age. (Their age should be the number of years old they were on their last birthday.) 12. A range is given instead of one number A. Note Function Key: Press to get into command mode, then hit the letter "n" and hit B. Type in the numbers of the range followed by 2 forward slashes (e.g. 15/20// ) C. Hit . Then you will need to leave a second note: D. Note Function Key: Press to get into command mode, then hit the letter "n" and hit E. Type 'Range//' F. Hit G. Enter the midpoint (see editor's marking) as the answer to the question H. Hit and move on to the next question NOTE: If the midpoint is NOT a whole number, round UP on odd numbers and DOWN on even numbers. For example, with a range of 15-20 minutes the midpoint is 17.5 so the answer would be 18 minutes. For a range of 10-15 years, the midpoint is 12.5 so the answer would be 12 years. 13. For the following series of questions (Page 7 if male and page 6 if female), 16. Has a medical professional ever said you have any of the illnesses or conditions listed below? 16a. How old were you when first diagnosed with this illness or condition? 16b. How much does each of your illnesses or conditions currently interfere with what you like to do? If the respondent answered 16 but not 16a/16b or answered 16a/16b but not 16, Enter a " Blank" for the unanswered questions 14. My immediate boss or supervisor / Other people at work (Questions 7 - 10, page 20 if female or page 18 if male) In the past, when the respondent circled 5 for the first item, the instrument would not allow an answer of 5 on the next three items. Please note that the instrument ihas been changed and now allows an answer of 5 for all of these items. The instrument is now properly behaving here. 15. When a respondent has scratched out an answer If the respondent answered one way, then scratched it out and answered another way, or if they scratched out their answer and did not fill out another one, you will leave an F1 note: This is a standard note--always type: 'Scratch//'. You will then need to leave a second F1 note that indicates what the respondent did. You will use your own words here, but be brief and try to follow this pattern: 'R scratched out Yes and marked No.' Another example would be 'R scratched out Yes but did not mark another answer.' If the respondent scratched out their answer and did not fill out another one, then enter for blank at that item. 16. The time to complete and comments on the last page A. We will be entering the time to complete and the respondent's comments written at the end of the questionnaire) B. Enter the comments verbatim. The field available for entry here is supposed to be huge (pages and pages) so we should not have a problem with long comments. If there happens to be not enough room, leave an F1 note stating this. C. If there are enclosures or attached letters, leave a note using the note function key (no need to type up the enclosure or letter) 17. If you must exit a case that you are not finished entering: Use the F4 function key, which will give you a screen that allows you to effortlessly exit the case. If you were in the middle of your first pass of the case, then you will have to go back into the case by entering the case id and then be sure to select 'first pass', so that you can continue where you left off. To continue exactly where you left off, you can use the F3 function key to 'check forward' which essentially checks through the data to take you to the last item that you entered. 18. Coding first or second pass: Be very careful when coding that a case has been entered first pass or entered second pass. This is not exactly intuitive because coding a case first pass completed means coding a '2'. Coding a case second pass completed means coding the case a '1'. Think of it as '1' means complete. Some protections have been put in place to ensure that the same person cannot do both the first and second passes on the same case. When in doubt, always check with a supervisor at this screen. Mistakes here can cause significant problems at the delivery end. It is possible to code a case second pass when only a first pass has been complete--so it is your responsibility to get this right, as no protection exists here. 19. When finalizing a case: Finalize a case as complete with problems, only if there were major problems that you were not able to resolve with a supervisor. When you finalize a case with problems, it must be filed in the Problem cases file drawer. 20. Before filing an entered survey: Data entry staff must handwrite either 1 (circled) for first pass, or 2 (circled) for second pass on the bottom left-hand corner of the back cover of the survey. Below the circled number, data entry staff should write their initials, and the full date (mm/dd/yyyy). 21. R doesn't give a definite answer: If the R answers something like 'approx 4,' it is O.K. to code the answer '4' and leave an F1 Note saying, R wrote 'approx 4'. Use this example as a guide. Updated 3/5/2004 K.D. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- F. Additional References and Resources -------------------------------------------------------------------------------- Wisconsin Longitudinal Study (WLS) MEMOS: 139 Notes on the Anxiety Symptoms & Beliefs Scale (Used in 1994 Sibling Mail survey) 140 WHO-Like Questions and Vignettes 141 Anxiety Measures on 2003 Mail Questionnaire 142 Brief Personality Measures for Mail Surveys 143 Reliability of Brandtstadter FGA and TGA Scales 144 Spielberger's Anxiety and Anger scales in 1994 Sibling Mail Questionnaire (Previously called Hostility) 145 Revised Plan for Implementing WHO Health Vignettes 146 Childhood Abuse items for the 2002 round and supporting documentation. 147 Reliability of Ryff's Model of Psychological Well-Being. Analytic memo describing choice of items for the 2002 round. 147c discusses the conceptual overlap of Ryff's model with other scales and shows why we dropped Pearlin-Schooler's Mastery Scale. Wisconsin Longitudinal Study (WLS) CORS: 764 Recoding BMI in 1992 graduate mail surveys and 1994 sibling mail surveys. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- J.Yonker 04/22/2005 -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- Added 07/17/2008 by Vicky Chang (based on cor571a.asc Appendix G) Family stress spillover to work index= q5wf_d + q5wf_j Work stress spillover to family index= q5wf_b + q5wf_g Family enhancement spillover to work index= q5wf_a + q5wf_e + q5wf_f Items were left as raw and then recoded before being summed for creation of the scales (1=5,2=4,3=3,4=2,5=1). This was done in order to have scales where a higher score indicates MORE of the construct--i.e., MORE FAMILY STRESS SPILL OVER TO WORK, MORE WORK STRESS SPILLOVER TO FAMILY, MORE FAMILY ENHANCEMENT SPILLOVER TO WORK. Source: MacArthur Research Network on Successful Midlife Development (MIDMAC) If you would like to learn more about the factor structure for these scales (a more expansive version also used in the MIDUS survey) see Grzywacz & Marks, 2000, Reconceptualizing the work-family interface..., Journal of Occupational Health Psychology, 5, 111-126. -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- -------------------------------------------------------------------------------- End of Document.