REs 2`12 Research wk 9 displaying the data

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/60

encourage image

There's no tags or description

Looks like no tags are added yet.

61 Terms

1
New cards

Does 'data' only mean numbers?

No. Data can be numerical or non-numerical, and must always be examined within a CONTEXT

2
New cards

Why must data always be examined within a context?

Without context (what the values mean, who they describe), a table of raw values tells us nothing

3
New cards

What does 'What' mean in the W's (and H) of data?

What information is being collected: the characteristics being recorded, counted, or measured

4
New cards

What are the characteristics being recorded in a data set called?

Variables

5
New cards

In a data table, what do the columns usually represent?

The variables (the 'what')

6
New cards

What does 'Who' mean in the W's (and H) of data?

Who or what is being studied: the individuals, subjects, objects, or events about which data are collected

7
New cards

In a data table, what do the rows usually represent?

The individuals/cases (the 'who')

8
New cards

What does 'How' mean in the W's (and H) of data?

How the data were collected: the methods and research design used to obtain the information

9
New cards

Why does 'How' the data were collected matter?

It helps us evaluate the quality and reliability of the data

10
New cards

In the Amazon vinyl purchase table, what is the 'who' (what does each row represent)?

One order/purchase of a vinyl album

11
New cards

In the Amazon vinyl purchase table, name the variables (columns).

Order #, Province, Price, Album Name, Area Code, Gift?, ASIN, Name

12
New cards

What are the four types of 'what' (variables) in this unit?

Qualitative (categorical), Quantitative, Identifier, Ordinal

13
New cards

What is a qualitative (categorical) variable?

A variable that describes characteristics or places observations into categories; can be numerical or non-numerical

14
New cards

Examples of qualitative (categorical) variables?

Province, blood type, smoking status

15
New cards

What is a quantitative variable?

A measured numerical value with measurement units; represents measurable or countable quantities

16
New cards

Examples of quantitative variables?

Age, respiratory rate (RR), number of hospital admissions

17
New cards

What is an identifier variable?

A variable used solely to uniquely identify each case; may contain letters or numbers but does not represent quantities

18
New cards

Examples of identifier variables?

Patient ID, order number

19
New cards

What is an ordinal variable?

A type of categorical variable with a meaningful order or ranking, where differences between categories are not necessarily equal

20
New cards

Examples of ordinal variables?

Pain severity, satisfaction ratings

21
New cards

Area code 819 appears as a number in the vinyl table. What type of variable is it, and why?

Categorical, because it names a location category and does not represent a measurable quantity

22
New cards

Is the order number (e.g., 105-36888) quantitative?

No. It is an identifier: it only uniquely identifies each case and does not represent a quantity

23
New cards

What are the 3 rules of data analysis?

Make a picture, make a picture, make a picture!

24
New cards

Why do we display data? (THINK, SHOW, TELL)

To THINK clearly (reveal things you may not see), SHOW important features and patterns, and TELL others about your data

25
New cards

What are common methods of displaying qualitative data?

Bar graphs, frequency tables, pie charts

26
New cards

What requirement applies to the type of display chosen?

It should accurately represent the data without misleading the reader

27
New cards

Which number always carries the percentage?

The denominator (the total number of data points or participants it is based on)

28
New cards

Why does a percentage alone not tell the whole story?

It does not tell us how many observations were used to calculate it

29
New cards

Study A: 2/4 have arthritis; Study B: 50/100. Both report 50%. Why is this a concern?

The percentages match, but Study A's sample is tiny; one more observation would change its percentage a lot

30
New cards

Why can small samples produce misleading percentages?

A single additional observation can change the percentage substantially

31
New cards

Which sample generally gives a more precise estimate: 4 people or 100 people?

The larger sample (100), assuming appropriate sampling methods

32
New cards

How should percentages be reported?

Alongside the actual counts and total sample size whenever possible

33
New cards

What is a frequency table?

A table listing each category and the count of observations in it (with a total)

34
New cards

In the frequency table (First 325, Second 285, Third 706, Crew 885), what is the total?

2201

35
New cards

What is a relative frequency table?

A frequency table that also shows each category's percentage of the total

36
New cards

How do you calculate a percentage in a relative frequency table? (heart transplant example: 3 of 86)

Frequency / total x 100, e.g., 3/86 = 0.0349 x 100 = 3.49%

37
New cards

What was the heart transplant data set about?

86 patients referred for heart transplant, with age grouped into decade bands to compare ages

38
New cards

Which decade band had the largest percentage of heart transplant referrals?

40-49 (21 patients, 24.42%)

39
New cards

Why can the '60 and over' (>=60) category be misleading?

We don't know the maximum age, so the band's range is unclear

40
New cards

What is the area principle?

The area occupied by part of a graph should correspond to the magnitude of the value it represents

41
New cards

In a bar graph, what should change between categories to follow the area principle?

Only the height of the bars (width should stay the same)

42
New cards

Where should the vertical axis of a bar graph begin?

At zero, to accurately represent differences between categories

43
New cards

What scale rule applies to graphs?

They should use a consistent scale

44
New cards

A bar graph's axis doesn't go all the way to 100%. What should you do?

Check the axis; don't assume the scale

45
New cards

What is a contingency table?

A table that explores the relationship between two categorical (qualitative) variables

46
New cards

What does a contingency table show?

How individuals are distributed along each variable, contingent on the value of the other variable(s)

47
New cards

What is the advantage of a contingency table?

It allows us to compare values more fairly

48
New cards

In the heart transplant contingency table, which decade band had the most survivors?

40-49 (17 alive)

49
New cards

Does the 40s band having the most survivors mean they have the best survival rate?

No. We must look at the data comparatively, as percentages of each band's total

50
New cards

What is a conditional distribution?

The distribution of one variable (e.g., outcome) within a subgroup defined by a condition of another (e.g., an age band)

51
New cards

How is survival rate within a decade band calculated? (20-29 example)

Alive / total in that band x 100: 8/9 = 0.8889 x 100 = 88.89%

52
New cards

Which decade bands had a 100% survival rate?

0-9 and 10-19

53
New cards

What happened to survival rates as age increased?

Younger bands had the highest survival rates; survival dropped to 50% for ages 50-59 and 33.33% for >60

54
New cards

Can we conclude from the survival table that age causes survival differences after heart transplant?

No. We cannot definitively say the pattern is related to the heart transplant itself

55
New cards

Percentage of total vs percentage of column: which shows each decade band's survival rate?

Percentage of each column (e.g., 17/21 = 80.95% for the 40s)

56
New cards

What does '% of total' show in the heart transplant table?

Each cell as a share of all 86 patients (e.g., 40s alive: 17/86 = 19.76%)

57
New cards

Compare: A) % of referrals in their 40s who survived AND B) % of 40s referrals who survived.

A = 17/86 = 19.76% (of all patients); B = 17/21 = 80.95% (of the 40s band only)

58
New cards

What should you ask when you look at a chart or display?

Is the area principle being violated? Be critical of how it is presented

59
New cards

Why is caution needed with similar-sounding percentages?

They can use different denominators and mean very different things

60
New cards

What do observed differences in a display suggest, and what is needed next?

They may suggest an association, but statistical testing is needed to evaluate the evidence

61
New cards