| Course | PHLT 8032 SPSS Revealed |
|---|---|
| Week | Week 2 |
| Paper type | Variable definition paper |
| Length | About 1,250 words, 7 pages |
| Format | APA 7 student paper |
| School | Walden University |
| Program | PhD in Public Health |
| Updated | October 2026 |
Free sample paper for PHLT 8032 Week 2
Variable Definitions for the Harlow County Adult Health Survey Extract: Measurement Levels Assigned and Argued, Derived Variables Named, and What the Level Setting Controls in SPSS
Student Name
Walden University
PHLT 8032: SPSS Revealed
Instructor Name
Month Day, Year
Variable Definitions for the Harlow County Adult Health Survey Extract: Measurement Levels Assigned and Argued, Derived Variables Named, and What the Level Setting Controls in SPSS
Purpose
Last week's setup gave each of the 16 variables in the composite county extract a name, a label and its missing-value codes, and left the Measure column provisional. This definition fixes that column. Each variable receives a measurement level, a one-phrase reason, and, where the choice is contestable, an argued entry stating what the level opens and what it leaves unchecked. Stevens (1946) proposed the nominal, ordinal, interval and ratio scheme on the grounds that the operations allowed on numbers depend on how the numbers were assigned. SPSS compresses interval and ratio into one setting called scale, so the file uses three levels. The definitions below describe respondents and are unweighted, and the two days items are treated as if code 88 had already been recoded to zero, since that recode is logged next week.
The Definition Table
Table 1
Measurement Levels Assigned to the Extract's Variables
| Variable | What it records | Stored type | Level | Reason in a phrase | |
|---|---|---|---|---|
| resp_id | Respondent identifier | String | Nominal | A label, never a quantity | |
| int_date | Interview date | Date | Scale | Dates subtract into intervals | |
| district | District of residence | Numeric code | Nominal | Places with no order | |
| age | Age in years | Numeric | Scale | Equal intervals and a true zero | |
| age_grp | Age group, four bands | Numeric code | Ordinal | Ordered bands of unequal width | |
| sex | Sex reported at interview | Numeric code | Nominal | Two categories, no order | |
| inc_cat | Household income bracket | Numeric code | Ordinal | Ordered, unequal brackets | |
| insured | Any health insurance | Numeric code | Nominal | Yes or no | |
| srh | Self-rated general health | Numeric code | Ordinal | Five ordered labels, 1 is best | |
| phys_days | Poor physical health days | Numeric | Scale | Counted days, 0 to 30 | |
| ment_days | Poor mental health days | Numeric | Scale | Counted days, 0 to 30 | |
| smoker | Smoking status, three groups | Numeric code | Nominal | Current, former, never | |
| height_in | Height in inches | Numeric | Scale | Measured quantity | |
| weight_lb | Weight in pounds | Numeric | Scale | Measured quantity | |
| bmi | Body mass index, computed | Numeric | Scale | Ratio of two quantities | |
| bmi_cat | BMI category, four bands | Numeric code | Ordinal | Ordered clinical cut points | |
| act_min | Weekly activity minutes | Numeric | Scale | Counted minutes | |
| flu_vax | Influenza vaccine, past year | Numeric code | Nominal | Yes or no | |
| wt_final | Final survey weight | Numeric | Scale | Not analyzed; applied | |
Derived variables (age_grp, bmi, bmi_cat) were computed from source variables that remain in the file; all levels are this analyst's assignments for the composite extract.
Entries That Needed an Argument
Self-Rated General Health
Self-rated health has five labeled categories: excellent, very good, good, fair and poor. It is defined as ordinal because the categories are ordered but nothing establishes that the step from excellent to very good equals the step from fair to poor. Its coding runs from 1 for excellent to 5 for poor, so a higher number means worse health. That direction matters for every later procedure: a positive correlation between income and this code would mean that richer respondents report worse health, the opposite of what a reader might assume. The entry therefore states the direction in words beside the level. Treating the rating as scale would permit a mean, and some analysts defend that practice when ratings are averaged across many items (Norman, 2010). A single item with five labels is a weaker case, and Jamieson (2004) argues that the mean of such an item is hard to interpret. This file keeps it ordinal and will report the share in fair or poor health rather than a mean rating.
Income Bracket
Income is collected in five brackets: under $25,000, $25,000 to $49,999, $50,000 to $74,999, $75,000 to $99,999, and $100,000 or more. The brackets are ordered, but their widths differ and the top one has no upper limit, so the codes 1 to 5 cannot be read as amounts. Ordinal is the only defensible level. The entry notes one consequence: a correlation computed on these codes as though they were dollars would weight the gap between brackets 4 and 5 the same as the gap between 1 and 2, which the data do not support.
Days of Poor Physical and Mental Health
The two days items count days in the past 30 when health was not good, following the Healthy Days measures the CDC developed for population surveillance (Moriarty et al., 2003). A count with a true zero and equal one-day steps is scale. The argument concerns what scale allows rather than whether it applies. Most respondents will report zero, and the counts are bounded at 30, so the distribution will be piled at one end. A scale setting permits a mean, but the mean of such a distribution may describe few respondents. The entry records that the level is scale while warning that medians, and the share reporting 14 or more days, will often be the better summary.
BMI and Its Category
Body mass index is computed from height and weight and is scale. The derived category, with bands at 18.5, 25 and 30, is ordinal. Both versions remain in the file because collapsing a measure into bands discards information that later procedures may want back. The entry also names the source of the derived version so any later table can say which one it used.
Age and Age Group
Age in years is scale. Age group, with bands 18 to 34, 35 to 49, 50 to 64 and 65 or older, is ordinal, and its bands are unequal, which is acceptable for an ordered grouping but not for arithmetic.
Smoking Status
Smoking status has three groups. It is tempting to call it ordinal, with never, former and current as increasing exposure, but former smokers vary widely in how long ago they quit and how much they smoked. Velleman and Wilkinson (1993) argue that a variable's level depends on the question being asked rather than on the data alone. For questions about current exposure, nominal treatment avoids implying an order the categories do not carry, and the entry says that a different question could justify a different choice.
What the Level Setting Controls in SPSS
The Measure column does not police the classic procedures. Frequencies, Descriptives and Means will compute a mean of district codes or income brackets without any warning, because they use whatever numbers they are given (Field, 2018). The setting matters in three places this course uses. The Chart Builder reads it to decide whether a variable becomes a category axis or a continuous one, so an income bracket left as scale would be drawn as a continuous axis with gaps. Custom Tables uses it to decide which statistics a variable can carry. Some newer dialogs, including the automated nonparametric tests, use it to choose a procedure. The practical lesson is that the definitions above are a record of reasoning for the analyst and the reader, not a safeguard. A mean income code will still appear if someone asks for it, which is why every later write-up will state which summary each variable receives and why.
Summary of Decisions
Six variables are nominal, four are ordinal and eight are scale, counting the three derived variables and leaving the weight outside the analysis. The two choices most likely to be questioned, self-rated health as ordinal and smoking status as nominal, are argued above with the alternative named. Next week's cleaning log begins from these definitions.
References
Field, A. (2018). Discovering statistics using IBM SPSS Statistics (5th ed.). SAGE.
Jamieson, S. (2004). Likert scales: How to (ab)use them. Medical Education, 38(12), 1217-1218. https://doi.org/10.1111/j.1365-2929.2004.02012.x
Moriarty, D. G., Zack, M. M., & Kobau, R. (2003). The Centers for Disease Control and Prevention's Healthy Days Measures: Population tracking of perceived physical and mental health over time. Health and Quality of Life Outcomes, 1, Article 37. https://doi.org/10.1186/1477-7525-1-37
Norman, G. (2010). Likert scales, levels of measurement and the "laws" of statistics. Advances in Health Sciences Education, 15(5), 625-632. https://doi.org/10.1007/s10459-010-9222-y
Stevens, S. S. (1946). On the theory of scales of measurement. Science, 103(2684), 677-680. https://doi.org/10.1126/science.103.2684.677
Velleman, P. F., & Wilkinson, L. (1993). Nominal, ordinal, interval, and ratio typologies are misleading. The American Statistician, 47(1), 65-72. https://doi.org/10.1080/00031305.1993.10475938
Reading the PHLT 8032 Week 2 instructions
The second assignment usually asks students to assign a measurement level to each variable in the course data file and to justify the assignment. A table of levels is the expected core, but faculty read the justifications most closely, especially for variables where reasonable analysts disagree. Strong submissions state the coding direction of any rating scale, explain why ordered brackets of unequal width cannot be treated as amounts, and say what a scale setting would permit that an ordinal one would not. Derived variables, such as a category computed from a continuous measure, belong in the table with their source named. Many prompts also ask what the level setting does in the software, which tests whether the student knows that most classic procedures will compute on any number regardless of the label.
How this PHLT 8032 Week 2 example is built
An opening paragraph links the definitions to last week's provisional Measure column and states that the scheme comes from Stevens. The definition table follows, one row per variable with what it records, its stored type, the level and a short reason, including three derived variables. Argued entries then take the hard cases in turn. Self-rated health is set ordinal, with its reversed coding direction spelled out and the case for averaging Likert items weighed. Income is ordinal because its brackets are unequal and open-ended. The days counts are scale but flagged for medians. BMI and its bands are both kept. Smoking status is defended as nominal, citing the argument that level depends on the question asked. A final section explains what the setting governs in the Chart Builder, Custom Tables and newer dialogs.
PHLT 8032 Week 2 rubric: what earns full marks
Most of the grade rides on the arguments, not the table. A correct list of levels with no reasons earns the base share, while entries that name the alternative and its consequence earn the rest. Graders check self-rated health closely, both for the level chosen and for whether the coding direction is stated, since a reversed scale changes the meaning of every later coefficient sign. Income treated as scale, or a mean income code proposed, loses marks. Derived variables need their source recorded. Credit also goes to recognizing that the Measure column is advisory for Frequencies and Descriptives but consequential in the Chart Builder. Sources on measurement theory should be cited for the contested choices rather than for definitions everyone knows, and the table should carry APA numbering with a note.
Common PHLT 8032 Week 2 mistakes, and how to avoid them
Definition papers lose marks when every variable gets the same one-word justification, such as "ordinal because it is ordered." That restates the level instead of arguing it. Give the easy cases a phrase and save the words for the variables that resist a label: rating scales, bracketed income, bounded counts and anything you computed yourself. For each, state what the chosen level allows, what it rules out, and what the other choice would have permitted. Say which way a scale is coded, in words, beside its level. Keep both versions of a variable you collapsed into categories, and record where the derived one came from. End with what the setting changes in the software. A definition table and argued entries built on your section's own variable list can be written to the same standard on request.
Get a PHLT 8032 Week 2 example written to your instructions
Attach the Week 2 prompt, its rubric and the variable list for the file your section assigned. A definition table with argued entries for the disputed variables comes back inside 24-48h, with no fee on the first. Its variables come from a composite county survey, so the levels argued there are reasoning to weigh, not settings to copy.
More PHLT 8032 papers and PhD in Public Health samples
- PHLT 8032 Week 1: SPSS Setup Walkthrough
- PHLT 8032 Week 3: Data Cleaning Log
- PHLT 8032 Week 4: Descriptive Statistics Write-Up
- PHLT 8032 Week 5: Chart Rationale Paper
- PHLT 8032 Week 6: Peer Replies on SPSS Output
- PHLT 8032 Week 7: Mann-Whitney Procedure Write-Up
- PHLT 8032 Week 8: Tau-b Association Run
- PHLT 8032 Week 9: APA Results Tables
- PHLT 8032 Week 10: SPSS Syntax Audit
- PHLT 8032 Week 11: Final Analysis Report
- PHLT 8006 Week 2: Research Habits Memo
- HLTH 9101 Week 2: Dissertation Argument Audit
- HLTH 8815 Week 2: Diagnosis or Function Post
- HLTH 8745 Week 2: Wound Healing Mechanism Brief
PHLT 8032 Week 2 questions, answered
Is self-rated health ever defined as a scale variable?
Some analysts treat it that way, particularly when it is summed with other items, but a single rating with five labeled categories is usually defined as ordinal in this week. The definition can note the debate and say what a scale setting would permit later, such as a mean. The rubric rewards a position argued through to its consequence, not a particular answer delivered bare.
Does the measurement setting change the results SPSS produces?
For most classic procedures, no. Frequencies, Descriptives and many tests compute whatever they are given, so a nominal code can be averaged without an error message. The setting matters mainly in the Chart Builder, custom tables and some newer dialogs, which use it to decide how a variable is handled. That is why the definition is argued in words: the software will not police it for you.
Should derived variables appear in the definition table?
Yes, with their source named. A BMI category computed from height and weight is a separate variable with its own level, and a later table has to say which version it uses. The definition also notes that the original scale version stays in the file, since collapsing a measure into bands discards information that some later procedures will want back.
Where can I find a free PHLT 8032 Week 2 sample paper?
The full Week 2 variable definition paper is reproduced on this page, with a 19-row table of levels, argued entries for the contested variables and a section on what the Measure column controls in SPSS.
Should smoking status be ordinal or nominal?
Either can be defended. This sample treats it as nominal because former smokers differ widely in exposure, and it notes that a question about cumulative exposure could justify an ordinal choice instead.