You're working as a sports journalist at a major online sports media company, specializing in soccer analysis and reporting. You've been watching both men's and women's international soccer matches for a number of years, and your gut instinct tells you that more goals are scored in women's international football matches than men's. This would make an interesting investigative article that your subscribers are bound to love, but you'll need to perform a valid statistical hypothesis test to be sure!
While scoping this project, you acknowledge that the sport has changed a lot over the years, and performances likely vary a lot depending on the tournament, so you decide to limit the data used in the analysis to only official FIFA World Cup matches (not including qualifiers) since 2002-01-01.
You create two datasets containing the results of every official men's and women's international football match since the 19th century, which you scraped from a reliable online source. This data is stored in two CSV files: women_results.csv and men_results.csv.
The question you are trying to determine the answer to is:
Are more goals scored in women's international soccer matches than men's?
You assume a 10% significance level, and use the following null and alternative hypotheses:
# Start your code here!
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import scipy.stats as statsmen=pd.read_csv("men_results.csv")
women=pd.read_csv("women_results.csv")#EDA
print(men.head())
print(men.describe())
print(men.isna().sum())
print(men.dtypes)
print(men["tournament"].value_counts().sum())
print(women.head())
print(women.isna().sum())
print(women.dtypes)
print(women["tournament"].value_counts().sum())#adding total score
men["total_score"]=men["home_score"]+men["away_score"]
women["total_score"]=women["home_score"]+women["away_score"]
#removing the rows that are not FIFA World Cup
men_filtered=men[men["tournament"]=="FIFA World Cup"]
women_filtered=women[women["tournament"]=="FIFA World Cup"]
#converting the date to date format and removing also the matches prior to 2002-01-01
men_filtered["date"]=pd.to_datetime(men_filtered["date"])
women_filtered["date"]=pd.to_datetime(women_filtered["date"])
men_filtered=men_filtered[men_filtered["date"]>="2002-01-01"]
women_filtered=women_filtered[women_filtered["date"]>="2002-01-01"]
print(men_filtered.head())fig,ax=plt.subplots(2,1,sharex=True)
ax[0].hist(data=men_filtered, x="total_score",bins=10)
ax[0].set_title('MEN',color="b")
ax[0].axvline(x =men_filtered["total_score"].mean() , color = 'g', label = 'mean')
ax[1].hist(data=women_filtered, x="total_score",bins=10, color="r")
ax[1].set_title('WOMEN', color="r")
ax[1].axvline(x =women_filtered["total_score"].mean() , color = 'g', label = 'mean')
plt.show()
As we can see, the distributions are highly right skewed. Therefore, we cannot asume normality of the data. Therefore, we will have to use a non-parametric hypothesis test.
men_filtered['category'] = 'masculine'
women_filtered["category"] = 'feminine'
both = pd.concat((men_filtered, women_filtered))
sns.boxplot(x="total_score",y="category", hue="category",data=both)
plt.show()#Hypothesis test
#Because there are two independent groups, men's and women's, this scenario requires an unpaired two-sample test.
#since the data are not normal, we cannot use unpaired t-test. Therefore, Wilcoxon-Mann-Whitney will be the adequate test to use.
#since the alternative hypothesis is that women score more than men, we will use a one-tail test.
p_val=stats.mannwhitneyu(women_filtered["total_score"], men_filtered["total_score"], alternative="greater").pvalue
alpha = 0.01 #as required
if p_val < alpha:
result = "reject"
else:
result = "fail to reject"
# Store the result in a dictionary
result_dict = {"p_val": p_val, "result": result}
# Print or use the result_dict as needed
print(result_dict ) **Since the nil hypothesis was rejected, we can conclude that women matches have a greater total score on average. **