MATH 105 WAN Directed Reading 5 Flashcards
In the first two paragraphs of Understanding Types of Variables, the author points out that information about types of variables affects your choice of _________________. The type of calculations that make sense depend on variable type. With some variables it only makes sense for one case (or individual or subject) to have one response while with other variables multiple responses per case make sense. Even the choice of tools (table/graphs) for
_____________________________ is affected by the variable and number of responses.
Comparisons/Presenting Numbers
____________variables take on values that take on negative, zero, and positive values. They can be compared using the mathematical operation of _______________________ but cannot be compared using __________________.
Summarize the temperature example, indicating the comparison that is and is not appropriate.
Interval/Subtraction/Division
_______________ variables can be compared using _______________
______ and ____________________ because:
Summarize the distance example showing that both comparisons are appropriate.
Ratio/Subtraction/Division
A value of zero can be interpreted as the lowest possible value
Ratio variables can be interpreted two ways; if A is 2 miles from B but C is 4 miles from B, C is 2 miles farther than, or twice as far as, A
Categorical variables are of two types. ____________ variables are those where the categories have order. The three examples provided for this type of variable are:
____________ variables are those where the categories have no inherent order, such as:
Ordinal (Ordered)
Letter grades, Income grouped into ranges of several thousand dollars (units must be specified), Likert-type items
Nominal (Named)
Gender, Race, Religion
Summarize the paragraph on regrouping continuous/quantitative variables into categorical variables. Use the age example in your summary.
Categorical variables of continuous variables, like grouping age into 5 or 10 year age groups, can simplify information, or indicate whether values reach a cutoff like retirement age
When each case/individual can have only one valid value for a categorical variable, then every case/ individual should fall into: _____________________________________________
__________.
Mutually exclusive categories are ________________________.
Exhaustive categories _________________________________________
___________________.
Explain why the categories “under 18 years”, “18‐64” and “65 and older” are mutually exclusive and exhaustive to someone that does not understand those terms.
One group/Nonoverlapping/Encompassing all possible responses
These categories are mutually exclusive, because someone can only be one age at a given time, and exhaustive, because everyone has an age in one of these categories
Discuss the Census Bureau problems (and their solutions) that were finally addressed in 2000.
There were questions on what race you were, but some people considered themselves biracial, so the answers weren’t mutually exclusive until multi-race categories were implemented. There were also questions that treated hispanic race and hispanic origin as different characteristics, so for people who believed they were the same, the questions weren’t exhaustive
What are some ways than “other” categories are used?
Sometimes the “other” category misses out important information. What can be done in the data collection process to avoid this problem?
To allow for answers researchers didn’t anticipate, or to allow researchers to later combine uncommon responses instead of creating separate categories for reasons mentioned by only a small share of respondents.
Allow respondents to specify what “other” means in their case
In the health insurance example, a “family” is a case/individual. Summarize this example to illustrate how a variable might have no answer, only one answer or multiple answers depending on the situation of any specific case/individual.
In situations where answers can
vary from none to many, the categories are not ___________________ or
_________________________. In the situation where many individuals had multiple responses, when you add up all the responses, they can _______________ the numbers of case/individuals. If none of the responses apply for some individuals, it is possible that the number of responses could be __________________ the number cases/individuals.
A family can have the same health insurance for everyone, different health insurance for different people, or no health insurance at all, depending on the family’s situation.
Mutually Exclusive/Exhaustive/Exceed/Less Than
Self‐contained Tables (as well as charts/graphs) are those that are labeled so the audience can:
Using the title, row and column headings and notes, the audience should be able to discern (list all):
Understand the information without reference to the text
The purpose of the table, the context of the data (the W’s), the location of specific variables within the table, units of measurement or categories for every number in the table, data sources, definitions or pertinent terms and abbreviations
The title of each table (chart/graph) should convey:
The specific topics or questions addressed in that table
If there are multiple tables/charts/graphs, the individualized titles should:
Differentiate the titles from one another and to convey where each fits in the overall scheme of your analysis
Summarize the information in each of the following paragraphs:
a) Topic
b) Context
c) Units
Include the major components of the relationships illustrated in that table, and use summary phrases or name broad conceptual categories
Specify the context of the data by listing the W’s in the table title; where and when the data was collected, and the restrictions on who is included in the data. Name institution or name of specific study if necessary
State the units of measurement, level of aggregation, and system of measurement for every variable in the table. Generalize units for the table, and if the same units apply to most numbers in the table, specify in the title, or the column or row headings if there’s not enough space
Copy the “best” title, then identify the part of that title that indicates the topic, then the parts that give context.
Means and Standard Deviations for Soil Components, 100 Study Sites, Smith County, 1990
Topic: Means and standard deviations
Context: 100 study sites, Smith County, 1990
If units of measurement differ across rows/columns, mention them in the labels. Different variables often have different units. COMPLETE: Do not assume that the units of measurement will be self‐evident once the concepts are named:
Without labels, readers might erroneously presume that age was measured in months or years, or weight or length reported in British rather than metric units