Scribd
Upload
10 views3 pages

Part II Assignment

_Part_II_Assignment for a biostatistician position interview process

Uploaded by

Moustafa Kardjadj
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
10 views3 pages

Part II Assignment

_Part_II_Assignment for a biostatistician position interview process

Uploaded by

Moustafa Kardjadj
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Assignment Part II

The present report provide a general descriptive summary and EDA of the laboratory data.Multinational
logistic regression model was preformed to explore potential evidence that the IP had any effect on the
lab parameters through the course of the study.

EDA :
Missing Values :
## # A tibble: 2 × 4
## variables types missing_count missing_percent
## <chr> <chr> <int> <dbl>
## 1 LR_FGLUC numeric 15 62.5
## 2 LR_RBC numeric 1 4.17

Visualize data elements containing missing values

Outliers, Negative and Zerro values:


## # A tibble: 5 × 10
## variables min Q1 mean median Q3 max zero minus outlier
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <int> <int> <int>
## 1 LR_RBC 3.99 4.68 4.89 4.82 5.13 5.55 0 0 1
## 2 LR_MCH 27.9 28.7 32.7 29.4 31.0 100 0 0 1
## 3 LR_BASO 0 0.1 0.0792 0.1 0.1 0.1 5 0 5
## 4 LR_GRAN 0 0 0 0 0 0 24 0 0
## 5 LR_NRBC 0 0 8.38 0 0 111 22 0 2

Visualize data element with an outlier ratio of 5% or higher.


Multinational logistic regression model:
PS:I couldn’t find an R package for Multinational logistic regression model with repeated measurement
(With a fixed effect) within 2-4 hours assignment time, therefore, I have preformed a simple
Multinational logistic regression model.
Multinational logistic regression model (Package nnet of R) was preformed to explore potential evidence
that the IP (studyID) had any effect on the lab parameters (covariates) through the course of the study.
Covariates listed above with missing, outliers and zero values were excluded from the model.
No statistical effect at 95% confidence interval was found when “EFG-058” and “ABC-007” were chosen
as a response reference level for the model. A statistical significant impact of IP on all lab parameters was
found where “LMN-045” was chosen as a response reference level.
## # weights: 39 (24 variable)
## initial value 25.268083
## iter 10 value 1.291536
## iter 20 value 0.010585
## iter 30 value 0.000661
## final value 0.000000
## converged

## Call:
## multinom(formula = STUDYID ~ LR_WBC + LR_RBC + +LR_Hb + LR_Hct +
## LR_MCV + LR_RDW + LR_MCHC + LR_PLT + LR_MONO + LR_NEUT +
## LR_LYMPH, data = exo)
##
## Coefficients:
## (Intercept) LR_WBC LR_RBC LR_Hb LR_Hct LR_MCV LR_RDW
## ABC-007 2.341557 27.90549 24.80390 2.544645 1.019328 -6.70449 -0.8498743
## EFG-058 -4.134831 34.98933 -43.11275 5.594730 -1.813839 -2.45378 4.1115172
## LR_MCHC LR_PLT LR_MONO LR_NEUT LR_LYMPH
## ABC-007 -0.9082509 0.2651219 -3.355211 26.29262 20.19460
## EFG-058 -2.1659576 0.3806139 13.200324 -29.62092 17.83826
##
## Std. Errors:
## (Intercept) LR_WBC LR_RBC LR_Hb LR_Hct LR_MCV
## ABC-007 0.3045642 2.345144 1.559369 48.12114 0.1400995 27.41078
## EFG-058 8.9776563 18.559136 79.185022 1506.12988 5.7606314 581.49072
## LR_RDW LR_MCHC LR_PLT LR_MONO LR_NEUT LR_LYMPH
## ABC-007 3.989791 104.4655 99.59249 0.2131949 1.18780 0.7614105
## EFG-058 85.540061 2088.1069 3118.25984 0.2194806 25.67926 2.7171507
##
## Residual Deviance: 3.15809e-08
## AIC: 48

## (Intercept) LR_WBC LR_RBC LR_Hb LR_Hct LR_MCV


## ABC-007 1.487699e-14 0.00000000 0.0000000 0.9578275 3.446132e-13 0.8067714
## EFG-058 6.451078e-01 0.05939086 0.5861278 0.9970362 7.528618e-01 0.9966331
## LR_RDW LR_MCHC LR_PLT LR_MONO LR_NEUT LR_LYMPH
## ABC-007 0.8313174 0.9930631 0.9978760 0 0.000000 0.000000e+00
## EFG-058 0.9616641 0.9991724 0.9999026 0 0.248707 5.201217e-11

You might also like