In Stat-I, we described data by three different ways. Qualitative vs Quantitative Discrete vs Continuous Measurement Scales Describing Data Types.

In Stat-I, we described data by three different ways.

Qualitative vs QuantitativeDiscrete vs ContinuousMeasurement Scales

Describing Data Types

Qualitative Data - Sometimes referred to as Attribute or Categorical Data.Describes a non-numeric characteristic.Examples -

Poor, Fair, ExcellentRed, Blue, GreenShort, Medium, TallMale, FemaleGroup One, Group Two, Group Three,

etc

Qualitative vs Quantitative

Quantitative Data is something that can be quantified,that is to say, something that can can becounted or measured.

Discrete Data represent countable items. Continuous Data usually apply to measurements.

To quantify qualitative data - apply a number scale.Example #1: Poor Fair Excellent

1 3 5

Example #2: Female = 1 Male = 2

Quantitative Data

Nominal - Name only (arbitrary)Examples: Area Codes, ZIP Codes, Sports Jerseys

Ordinal - Order (but no defined interval)Example: Horse race - 1st, 2nd, 3rd, etc

Interval - Equal IntervalsExamples: Thermometer, Meter Stick, Speedometer

Ratio - Absolute ZeroExamples: Celsius Scale has negative values.Yardstick and weight scales have absolute zero.

Scales of Measurement

JMP uses two somewhat differing categories.

Data Types Modeling Types Numeric Continuous Character Ordinal Row Nominal

Note the possible confusion with our previous definitions.

JMP Data and Modeling Types

Numeric Data refers to quantitative data (numbers),may be discrete or continuous values.JMP treats all numeric data as continuous.

Character Data applies to alphanumeric text.If classified as character data, then “numbers” aretreated as text characters.

Row Data applies to row characteristics.Affects appearance of graphical displays.We will not be concerned with row data.

JMP Data Types

Continuous refers to data measurements.Must be numeric data type.Used in arithmetic calculations.

Ordinal refers to discrete categorical data.May be either numeric or character data type.If numeric, the order is the numeric magnitude.If character, the order is the sorting sequence.

Nominal refers to discrete categorical data.May be either numeric or character data type.Treated as discrete values without implicit order.

JMP Modeling Types

As if the foregoing was not confusing enough,we also have to deal with Modeling Platforms.

The Modeling Platforms are used for statistical analyses.

Depending on the platform model, JMP uses different algorithms and sets of assumptions to arrive at the final calculated results.

JMP Modeling (Analysis) Platforms

Response Models Factors Models (Y Variable) (X Variable)Continuous Response Continuous FactorsNominal Response Nominal FactorsOrdinal Response Ordinal Factors

Analysis Models

Distribution of Y (Univariate)Fit Y by XMatched PairsFit ModelNon-Linear FitNeural NetsTime SeriesCorrelation (Bivariate & Multivariate)Survival & Reliability

Analysis Platforms

Univariate (One Variable)DistributionsHistogramsScatterplotsNormality TestingOne Sample Hypothesis Testing

Distribution of Y

Bivariate (Two Variables)Scatterplot with Regression CurveOne Way ANOVAContingency Table AnalysisLogistic Regression

Fit Y by X

For Fit Y by X

XContinuous Nominal

Continuous B i v a r i a t e t - T e s t s

S c a t t e r P l o t M e a n s

R e g r e s s i o n L i n e O n e - W a y A N O V A

Y L i n e F i t t i n g C o m p a r i s o n T e s t s

N o n - P a r a m e t r i c T e s t s

P o w e r s T e s t i n g

L S N & L S V

Nominal L o g i s t i c R e g r e s s i o n C o n t i n g e n c y T a b l e

C r o s s T a b s

The roles of X and Y (nominal & continuous)

determine the type of analysis.

Paired t - test

Matched Pairs

General Linear Models Multiple Regression Two and Three Way ANOVA’s Analysis of Covariance Fixed and Random Effects Nested and Repeated Measures

Fit Model

Requires user generated predictor equation, using iterative procedures.

Non-Linear Fit

Implements and analyzes standard types of neural networks.

Neural Nets

Analyzes univariate time series taken over equally spaced time periods.

Plots autocorrelations

Fits ARIMA and Seasonal (Cyclic) ARIMA’s

Incorporates smoothing models

Times Series

Bivariate and MultivariateScatterplot MatricesMultivariate OutliersPrinciple Components

Correlations

Models time until an event.Used in - Reliability Engineering Survival Analysis

Survival & Reliability