Les recrutés à l’heure sont-ils mieux payés ?

Dans le graphique que vous avez réalisé, les barres étaient presque égales. Cela va dans le sens de l’hypothèse qu’une variable omise — le niveau de poste — explique la différence de rémunération entre les nouvelles recrues et les employés actuels. Toutefois, le graphique montre une petite différence dans les salaires moyens des employés payés à l’heure. Testez s’il existe une différence de rémunération significative entre les nouvelles recrues payées à l’heure et les employés actuels payés à l’heure.

Cet exercice fait partie du cours

<cours>Analytique RH : explorer les données des employés avec R</cours>

Exercice interactif pratique

Essayez cet exercice en complétant ce code d’exemple.

# Load the package
___

# Filter the data to include only hourly employees
pay_filter <- ___

Modifier et exécuter le code

Cet exercice fait partie du cours

<cours>Analytique RH : explorer les données des employés avec R</cours>

IntermédiaireNiveau de compétence

4.8+

Commencer le cours gratuitement

In this chapter, you will get an introduction to how data science is used in a human resources context. Then you will dive into a case study where you'll analyze and visualize recruiting data to determine which source of new candidates ultimately produces the best new hires. The dataset you'll use in this and the other chapters in this course is synthetic, to maintain the privacy of actual employees.

Exercise 1: Welcome to the course!Exercise 2: Applications of human resources (HR) analytics Exercise 3: Looking at the recruiting data Exercise 4: Recruiting and quality of hire Exercise 5: Identifying groups in data Exercise 6: Sales numbers by recruiting source Exercise 7: Attrition rates by recruiting source Exercise 8: Visualizing the recruiting data Exercise 9: Visualizing the sales performance differences Exercise 10: Visualizing the attrition differences Exercise 11: Drawing conclusions

Gallup defines engaged employees as those who are involved in, enthusiastic about and committed to their work and workplace. There is disagreement about the strength of the connection between employee engagement and business outcomes, but the idea is that employees that are more engaged will be more productive and stay with the organization longer. In this chapter, you'll look into potential reasons that one department's engagement scores are lower than the rest.

Exercise 1: Analyzing employee engagement Exercise 2: Importing the survey data Exercise 3: Which department has the lowest engagement?Exercise 4: Comparing other factors by department Exercise 5: Visualizing the engagement data Exercise 6: Visualizing several variables at once Exercise 7: Visualizing several variables at once with facets Exercise 8: Drawing conclusions from graphs Exercise 9: Is that difference meaningful?Exercise 10: Statistical significance - disengaged employees Exercise 11: Statistical significance - vacation days Exercise 12: Drawing conclusions from the tests

When employers make a new hire, they must determine what the new employee will be paid. If the employer is not careful, the new hires can come in with a higher salary than the employees that currently work at the same job, which can cause employee turnover and dissatisfaction. In this chapter, you will check whether new hires are really getting paid more than current employees, and how to double-check your initial observations.

Exercise 1: Rémunérer les nouvelles recrues équitablement Exercise 2: Importer les données de rémunération Exercise 3: La différence est-elle significative ?Exercise 4: Biais d’omission de variables Exercise 5: Exemples de biais de variable omise Exercise 6: Quelles autres différences existent ?Exercise 7: À quoi ressemble désormais l’écart de rémunération ?Exercise 8: Les recrutés à l’heure sont-ils mieux payés ?

Exercice actuel

Exercise 9: Régression linéaire Exercise 10: Rémunération des nouvelles recrues : une régression simple Exercise 11: Rémunération des nouvelles recrues : prise en compte des niveaux de poste Exercise 12: Tirer des conclusions à partir des tests

Performance management helps an organization keep track of which employees are providing extra value, or below-average value, and compensating them accordingly. Whether performance is a rating or the result of a questionnaire, whether employees are rated each year or more often than that, the process is somewhat subjective. An organization should check that ratings are being given with regard to performance, and not individual managers' preferences, or even biases (conscious or subconscious).

Exercise 1: Joining HR data Exercise 2: Importing the two datasets Exercise 3: Joining performance data Exercise 4: Performance ratings and fairness Exercise 5: Focus on high performers Exercise 6: Comparing distributions of high performers Exercise 7: Checking for omitted variable bias Exercise 8: Logistic regression Exercise 9: Linear and logistic regression Exercise 10: Performance ratings: a simple logistic regression Exercise 11: Performance ratings: accounting for job levels Exercise 12: Conclusions and recommendations

In many industries, workplace safety is a critical consideration. Maintaining a safe workplace provides employees with confidence and reduces costs for workers' compensation and legal liabilities. In this chapter, you'll look for explanations for an increase in workplace accidents.

Exercise 1: Employee safety Exercise 2: Importing and joining the accident data Exercise 3: Where is the highest accident rate?Exercise 4: Where did the accident rate increase most?Exercise 5: Focusing on the location of interest Exercise 6: Focusing on the problem location Exercise 7: Bringing in more data Exercise 8: Checking for omitted variables Exercise 9: Is that change isolated to the problem location?Exercise 10: Explaining the increase in accidents Exercise 11: Using regression to identify change drivers Exercise 12: What can you conclude?Exercise 13: Conclusion