Data Analysis Chapter 1-2

Front 1

Basic format of a journal article

Back 1

1. Abstract

2. Introduction

3. Methods

4. Results

5. Discussion/Conclusion

Front 2

define abstract

Back 2

- summarizes article

- do not base worth of the article on the abstract

Front 3

Introduction

Back 3

1. Background - shows relationship between current research and previously published work

2. statement of purpose - the goal of the research

3. Hypotheses - may or may not be specifically stated

Front 4

Method

Back 4

1. Participants - description of subjects and how they were obtained

2. Materials/Instruments - things used to study in enough detail that it can be replicated. Reliability and validity included.

3. Procedures - describes what the subject did, where the study took place, the sequence of events and drop out of subjects.

Front 5

Where is the statistical plan located in a journal article

Back 5

the end of the methods section or beginning of results.

Front 6

Results

Back 6

- statistical information presented as text and/or tables and graphs often using statisitical notation

- most critical part of the research article, but often ignored because of lack of statistical knowledge

Front 7

Discussion/Conclusion

Back 7

- non-technical interpretation of the results

- explanation should focus on the results according to the original purpose of hypothesis

- resason may be given for the results of lack of agreement with the hypothesis

- suggestion for future research may be given

Front 8

References

Back 8

The appendix

- questionnaires

- visual and graphs

Front 9

Dependent Variable

Back 9

- the effects of unknown causes

- the even that needs explained in research

Front 10

Independent variable

Back 10

- suspected cause of the event (DV) usually not know

- the variable that is directly manipulated (treatment group)

Front 11

Research hypothesis

Back 11

what the research predicts concerning the relationship between the IDV and DV

Front 12

Null Hypothesis

Back 12

- statistical prediction concerning the RH but in the null sense

- research question is used if no prediction is made

Front 13

Research Problem

Back 13

Problem statements

- what the research is concerned with

Front 14

Level of measurement

Back 14

the relationship of the values that are assigned to the attributes for a variable

Front 15

why is level of measurement important?

Back 15

- helps you decided what statistical analysis is appropriate on the values that were assigned

- help you decide how to interpret the data from the variable

Front 16

what are the measurement levels

Back 16

1. Nominal

2. Ordinal

3. Interval

4. Ratio

Front 17

nominal

Back 17

-provides for classification

- no order, distance or origin

- race, region, name of condition

Front 18

ordinal

Back 18

- has order - indicating more or less of a certain quality

- no equal intervals between categories

- example: the amount of assistance needed to ambulate from maximal to independent rankings

- can be ranked but no equal distance

Front 19

example of ordinal

Back 19

0 = less than H.S.

1 = some H.S.

2 = H.S. degree

3 = some college

4 = college degree

5 = post college

Front 20

interval

Back 20

- real number properties of order and distance but no true origin in zero

- equal interval allow for addition and subtraction

- multiplication and division do not make sense because there is no true zero

- example: temps in fahrenheit

Front 21

ratio

Back 21

- order, distance and origin of 0

- + - X / are all possible

- example: length, weight, time, Kelvin temperature, ROM

Front 22

non-paremetric

Back 22

- nominal or ordinal

- ordinal includes ordered categories or ranks

Front 23

parametric

Back 23

- Quantity (Interval or Ratio)

Front 24

Range of scores

Back 24

tells us would this population be representative of my subjects

Front 25

Mean

Back 25

the "center" of the distribution of scores

Front 26

Standard Deviation (SD)

Back 26

the spread or variability of the scores

Front 27

P-value

Back 27

the probability of the scores being due to chance alone

Front 28

Qualitative vs. Quantitative

Back 28

- Qualitative = Race/Ethnic group

- Quantitative = number of deaths (discrete)

height and weight (continuous)

Front 29

frequency distribution

Back 29

a table that show the values a variable can take and number of observations of each value

- if more than 8 or 10 possible values, values are usually combined into class intervals

Front 30

examples of frequency distribution

Back 30

number of persons

age or age group

racial/ethnic group

time period

Front 31

stem and leaf displays

Back 31

- frequency distribution with no loss of information

- score intervals set up on left side (stem)

- shows all but last digit of each observation

- last digit is the leave of the plot

Front 32

histogram

Back 32

- graph of a frequency distribution

- columns are always right next to each other

- x-axis intervals should be a equal size

Front 33

a chart is a good for what?

Back 33

a good way to display with a qualitative aspect (nominal)

- examples: race/ethnicity

Front 34

what type of bar graph should be used when names are long?

Back 34

horizontal

Front 35

bar chart/graphs

Back 35

- horizontal or vertical

- vertical bar chart: x axis is a categorical variable or grouping of continuous

- bars are separated

Front 36

Pie Charts

Back 36

- simple and easily understood way to display relative proportions

- use with single variable or small groups of variables.

Front 37

what is the most useful for generally normal distribution

Back 37

the arithmetic mean

- average or "mean"

- most commonly used

- sum of all observations divided by number of obersvations

Front 38

median

Back 38

- middle of the set of data.

- if n is even, middle rank will be between two observations

- the mean of those two observations is the median

Front 39

mode

Back 39

the value with the most observations

- can be helpful when combined with median or mean , to describe the skewness of distribution

Front 40

normal distribution

Back 40

bell shaped curve

- mean, median and mode are the same in a normal distribution

Front 41

positive skew

Back 41

tail is towards positive

Front 42

negative skew

Back 42

tail points towards negative end

Front 43

mesokurtic curve

Back 43

baseline for comparison. Like normal curve

Front 44

leptokurtic curve

Back 44

high peak

Front 45

platykurtic curve

Back 45

low peak.

larger distribution of scores

Front 46

minimum and maximum value

Back 46

shows the boundaries of the data

- gives an idea of how spread out the data is

- max value - min value

Front 47

Percentiles

Back 47

Maximum is 100th percentile. 100% of value lie below the max

median is 50th percentile: 50% of values lie at or below the median

Front 48

box and whiskers plot

Back 48

- illustrate the variability of the data as well as the central tendency and the shape of the distribution

- line in box = median

- box = inter-quartile range representing 50% of scores

- whiskers are the smallest and largest values (5th and 95th percentile)

- values outside whiskers are outliers

Front 49

outliers

Back 49

scores that lie far away from the rest of the data, usually more than 3 SD.

can have interest to inspect data to determine reason for outliers.

Front 50

how much data lies within 1 SD above and below the mean?

Back 50

68%

Front 51

how much data lies within 2 SD above and below the mean?

Back 51

95%

Front 52

how much data lies within 3 SD above and below the mean?

Back 52

98%