2. Linear Model

The linear model is used to analyze the relationship between two or more variables.

Example 1:

The gain of a transistor is the difference between the emitter and the collector. The Gain variable (in hFE) can be controlled in the ion deposition process through the variables Emission time (in minutes) and Ion dose ($ \times 10^{14} $) variables. The data is shown in Table 1. Our aim is to evaluate the linear relationship between transistor gain and the covariates Emission Time and Ion Dose.


Time Dose Gain
195 4 1004
255 4 1636
195 4.6 852
255 4.6 1506
225 4.2 1272
225 4.1 1270
225 4.6 1269
195 4.3 903
255 4.3 1555
225 4 1260
225 4.7 1146
225 4.3 1276
225 4.72 1225
230 4.3 1321

We will upload the data to the system.

Configuring as shown in the figure below to perform the analysis.

Then click Calculate to get the results. You can also generate the analyses and download them in Word format.

The results are:

ANOVA Table

D.F. Sum of Squares Mean Square F Stat. P-value
Time 1 630967.86 630967.865 474.40781465 0
Dose 1 20998.23 20998.234 15.78800898 0.0026287995665368
Time:Dose 1 121.00 121.000 0.09097665 0.7691183430521420
Residuals 10 13300.12 1330.012

Expĺoratory Analysis (residues)

Minimum 1Q Median Mean 3Q Maximum
-44.58414 -25.94015 -3.266019 0 24.62883 63.20097

Coefficients

Estimate Standard Deviation T Stat. P-value
Intercept 71.1733231 1,970.46088 0.03612014 0.971897
Time 8.1533801 8.726180 0.93435849 0.372131
Dose -289.648874 457.471922 -0.6331511 0.540843
Time:Dose 0.6111111 2.026074 0.30162337 0.769118

Descriptive measure for Goodness-of-Fit

Standard deviation of residuals Degrees of Freedom R^2 Adjusted R^2
36.46932 10 0.980011 0.9740149

Confidence Interval for the parameters

2.5 % 97.5 %
Intercept -4319.287130 4461.633776
Time -11.289760 27.596520
Dose -1308.959837 729.662088
Time:Dose -3.903262 5.125484

Prediction Interval

Gain Time Dose Fitted Value Lower Limit Upper Limit Standard Deviation
1,004 195 4.00 979.1536 915.3027 1043.0045 28.656620
1636 255 4.00 1615.0231 1551.6616 1678.3846 28.436947
852 195 4.60 876.8643 815.6809 938.0477 27.459417
1506 255 4.60 1534.7338 1473.9136 1595.5539 27.296389
1272 225 4.20 1266.6586 1241.9861 1291.3311 11.073149
1270 225 4.10 1281.8735 1252.1883 1311.5586 13.322842
1269 225 4.60 1205.799 1174.5787 1237.0194 14.011854
903 195 4.30 928.0090 887.9554 968.0625 17.976226
1555 255 4.30 1574.8784 1535.4958 1614.2611 17.675135
1260 225 4.00 1297.0883 1261.0430 1333.1337 16.177353
1146 225 4.70 1190.5841 1152.7665 1228.4018 16.972751
1276 225 4.30 1251.4437 1229.4927 1273.3947 9.851714
1225 225 4.72 1187.5412 1148.3144 1226.7679 17.605162
1321 230 4.30 1305.3495 1282.8141 1327.8849 10.113995

Summary of Residuals Analysis

N.Obs Time Dose Residuals Studentized Residuals Standardized Residuals Leverage DFFITS DFBETA D-COOK
1 195 4.00 24.846387 1.1147927 1.1015027 0.61743967 1.41625490 -0.94199418165 0.4895597961
2 255 4.00 20.976913 0.9108462 0.9187053 0.60800973 1.13439076 0.75939078067 0.3272862194
3 195 4.60 -24.864288 -1.0402662 -1.0360193 0.56692711 -1.19022098 -0.75218753067 0.3512706755
4 255 4.60 -28.733762 -1.2162085 -1.1880774 0.56021533 -1.37266797 0.87352758220 0.4495152217
5 225 4.20 5.341425 0.1460048 0.1537206 0.09219065 0.04652787 -0.00004549144 0.0005999243
6 225 4.10 -11.873462 -0.3338476 -0.3497473 0.13345608 -0.13101533 0.00011212495 0.0047097360
7 225 4.60 63.200975 2.2127130 1.8770621 0.14761681 0.92082173 -0.00056023133 0.1525450911
8 195 4.30 -25.008951 -0.7720700 -0.7881510 0.24296384 -0.43739033 0.02076432661 0.0498406795
9 255 4.30 -19.878424 -0.6029964 -0.6231508 0.23489299 -0.33410935 -0.01574389071 0.0298039755
10 225 4.00 -37.088350 -1.1532997 -1.1347232 0.19677029 -0.57082381 0.00042262343 0.0788568872
11 225 4.70 -44.584138 -1.4566189 -1.3812096 0.21659532 -0.76590938 0.00035872582 0.1318627421
12 225 4.30 24.556312 0.6802984 0.6993418 0.07297401 0.19086997 -0.00019860744 0.0096248792
13 225 4.72 37.458840 1.1981057 1.1728416 0.23303688 0.66041999 -0.00029388986 0.1044886083
14 230 4.30 15.650523 0.4280325 0.4466624 0.07691128 0.12355207 0.00159139594 0.0041557123

Criterion

Diagnostic Formula Value
hii (Leverage) (2*(p+1))/n 0.57
DFFITS 2* Square root ((p+1)/n) 1.1
DCOOK 4/n 0.2857143
DFBETA 2/Square root(n) 0.53
Standardized Residuals (-3,3) 3
Studentized Residuals (-3,3) 3

Normality Test

Statistics P-value
Anderson-Darling 0.3592592 0.3978779
Shapiro-Wilk 0.9453507 0.4911092
Kolmogorov-Smirnov 0.1614289 0.4142572
Ryan-Joiner 0.9761196 0.5098

Homoscedasticity Test - Breusch Pagan

Statistics DF P-value
0.2119295 1 0.6452593

Homoscedasticity Test - Goldfeld Quandt

Variable Statistics DF1 DF2 P-value
Time 0.212367391304331 2 1 0.324428433008958
Dose 1.70140146940424 2 1 0.953159041390766

Independence Test - Durbin-Watson

Statistics P-value
1.71481 0.8201288

Lack of Fit Test

$\qquad \quad$Lack of fit test
Input data do not contain replicas

Outliers (X)

Observations Leverage Criterin
1 0.6174397 0.5714286
2 0.6080097 0.5714286

Outliers (atypical points)

Observation T-Value P-Value P-Value Bonferroni
7 2.212713 0.05420479 0.758867

Influential Point

Observations DFFITS Criterion
1 1.42 ± 1.07
2 1.13 ± 1.07
3 -1.19 ± 1.07
4 -1.37 ± 1.07

Influential Points

Observations DCOOK Criterio
1 0.4895598 0.2857143
2 0.3272862 0.2857143
3 0.3512707 0.2857143
4 0.4495152 0.2857143

Example 2:

The monthly net salary of employees at a federal university is linearly related to the variables City of employment (Campus the Campinas and Curitiba), Education (Higher Education Completed, Postgraduate Degree Completed or High School Degree Completed), Length of Service at the university, total bonuses in the month and Hours Worked in the month. Data was collected from 22 employees. This data can be analyzed in the table below.

Salary City Education Length of service years Bonus Hours worked
2024.84 Campinas ES 3 99 149.6
2017.15 Campinas POS 3 95.98 147.4
2014.94 Campinas EM 3 104.19 146.5
2047.4 Campinas POS 4 97.52 150.7
2043.59 Campinas POS 4 105.46 150.1
1913.15 Curitiba ES 4 94.83 140.2
2064.25 Campinas POS 5 99.19 150.9
1913.44 Curitiba ES 5 99.03 141.3
2076.47 Campinas ES 7 100.82 151.4
2075.83 Campinas ES 7 97.49 151.3
1950.89 Curitiba POS 7 107.05 142.8
1915.33 Curitiba POS 7 108.71 142.3
2096.7 Campinas ES 11 97.28 153.6
2092.04 Campinas ES 11 101.96 153.6
1988.75 Curitiba EM 11 106.77 143.6
1980.92 Curitiba ES 11 97.13 143.1
2115.13 Campinas ES 12 93.14 155.3
2101.45 Campinas ES 12 97.55 154.6
1999.28 Curitiba ES 12 106.29 145.6
2142.51 Campinas EM 13 101.89 156
2013.03 Curitiba EM 14 100.45 145.9
2232.56 Campinas EM 19 100.63 162.4

We will upload the data to the system.

Configuring as shown in the figure below to perform the analysis.

Then click Calculate to get the results. You can also generate the analyses and download them in Word format.

The results are:

ANOVA Table

D.F. Sum of Squares Mean Square F Stat. P-Value
City 1 76304.3862 76304.3862 587.6809 0.0000
Education 2 17610.2829 8805.1414 67.8154 0.0000
Length.of.service.Years 1 34129.9036 34129.9036 262.8616 0.0000
Bonus 1 71.7808 71.7808 0.5528 0.4686
Hours.worked 1 1445.4002 1445.4002 11.1322 0.0045
Residuals 15 1947.5974 129.8398

Exploratory Analysis (residues)


Minimum 1Q Median Mean 3Q Maximum
-23.2602 -6.455 0.1493 0 8.1092 13.7546

Coefficients


Estimate Standard Deviation T Stat. P-value
Intercept 814.3743 377.5276 2.1571 0.0476
CityCuritiba -45.6785 25.5988 -1.7844 0.0946
EducationES -16.1322 7.2616 -2.2216 0.0421
EducationPOS -13.7281 9.1100 -1.5069 0.1526
Length.of.service.Years 4.5184 2.1280 2.1233 0.0508
Bonus -0.5274 0.7177 -0.7348 0.4738
Hours.worked 8.4984 2.5471 3.3365 0.0045

Descriptive measure for Goodness-of-Fit


Standard deviation of residuals Degrees of Freedom R^2 Adjusted R^2
11.3947 15 0.9852 0.9793

Confidence Interval for the parameters


2.5 % 97.5 %
Intercept 9.6932 1619.0554
CityCuritiba -100.2411 8.8841
EducationES -31.6100 -0.6545
EducationPOS -33.1457 5.6894
Length.of.service.Years -0.0172 9.0541
Bonu -2.0570 1.0023
Hours.worked 3.0694 13.9274

Last modified 19.11.2025: Atualizar Manual (288ad71)