Inference for Numerical Data

ch7_inf_for_numerical_data

Inference for numerical data¶

North Carolina births¶

In 2004, the state of North Carolina released a large data set containing information on births recorded in this state. This data set is useful to researchers studying the relation between habits and practices of expectant mothers and the birth of their children. We will work with a random sample of observations from this data set.

Exploratory analysis¶

Load the nc data set into our notebook.

In [1]:
import warnings
warnings.filterwarnings("ignore")

import numpy as np
import pandas as pd
import io
import requests

df_url = 'https://raw.githubusercontent.com/akmand/datasets/master/openintro/nc.csv'
url_content = requests.get(df_url, verify=False).content


We have observations on 13 different variables, some categorical and some numerical. The meaning of each variable is as follows.

variable description
fage father's age in years.
mage mother's age in years.
mature maturity status of mother.
weeks length of pregnancy in weeks.
premie whether the birth was classified as premature (premie) or full-term.
visits number of hospital visits during pregnancy.
marital whether mother is married or not married at birth.
gained weight gained by mother during pregnancy in pounds.
weight weight of the baby at birth in pounds.
lowbirthweight whether baby was classified as low birthweight (low) or not (not low).
gender gender of the baby, female or male.
habit status of the mother as a nonsmoker or a smoker.
whitemom whether mom is white or not white.

Exercise 1

What are the cases in this data set? How many cases are there in our sample?

As a first step in the analysis, we should consider summaries of the data. This can be done using the describe() and info():

In [2]:
nc.describe()

Out[2]:
fage mage weeks visits gained weight
count 829.000000 1000.000000 998.000000 991.000000 973.000000 1000.00000
mean 30.255730 27.000000 38.334669 12.104945 30.325797 7.10100
std 6.763766 6.213583 2.931553 3.954934 14.241297 1.50886
min 14.000000 13.000000 20.000000 0.000000 0.000000 1.00000
25% 25.000000 22.000000 37.000000 10.000000 20.000000 6.38000
50% 30.000000 27.000000 39.000000 12.000000 30.000000 7.31000
75% 35.000000 32.000000 40.000000 15.000000 38.000000 8.06000
max 55.000000 50.000000 45.000000 30.000000 85.000000 11.75000
In [3]:
nc.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1000 entries, 0 to 999
Data columns (total 13 columns):
#   Column          Non-Null Count  Dtype
---  ------          --------------  -----
0   fage            829 non-null    float64
1   mage            1000 non-null   int64
2   mature          1000 non-null   object
3   weeks           998 non-null    float64
4   premie          998 non-null    object
5   visits          991 non-null    float64
6   marital         999 non-null    object
7   gained          973 non-null    float64
8   weight          1000 non-null   float64
9   lowbirthweight  1000 non-null   object
10  gender          1000 non-null   object
11  habit           999 non-null    object
12  whitemom        998 non-null    object
dtypes: float64(5), int64(1), object(7)
memory usage: 101.7+ KB


As you review the variable summaries, consider which variables are categorical and which are numerical. For numerical variables, are there outliers? If you aren't sure or want to take a closer look at the data, make a graph.

Consider the possible relationship between a mother's smoking habit and the weight of her baby. Plotting the data is a useful first step because it helps us quickly visualize trends, identify strong associations, and develop research questions.

Exercise 2

Make a side-by-side boxplot of habit and weight. What does the plot highlight about the relationship between these two variables?

The box plots show how the medians of the two distributions compare, but we can also compare the means of the distributions using the following function to split the weight variable into the habit groups, then take the mean of each using mean().

In [4]:
nc.groupby(['habit'])['weight'].mean()

Out[4]:
habit
nonsmoker    7.144273
smoker       6.828730
Name: weight, dtype: float64

There is an observed difference, but is this difference statistically significant? In order to answer this question we will conduct a hypothesis test .

Inference¶

Exercise 3

Check if the conditions necessary for inference are satisfied. Note that you will need to obtain sample sizes to check the conditions. You can compute the group size using the same groupby command above but replacing mean with size.

Exercise 4

Write the hypotheses for testing if the average weights of babies born to smoking and non-smoking mothers are different.

We will now conduct hypothesis tests for testing if the average weights of babies born to smoking and non-smoking mothers are different. For this task, we can use statsmodels, a Python module that provides classes and functions for the estimation of many different statistical models, as well as for conducting statistical tests, and statistical data exploration.

In [5]:
import statsmodels.stats.weightstats as st

nc_weightANDsmoker = nc[nc['habit'] == 'smoker']['weight']
nc_weightANDnonsmoker = nc[nc['habit'] == 'nonsmoker']['weight']

dsw1 = st.DescrStatsW(nc_weightANDsmoker)
dsw2 = st.DescrStatsW(nc_weightANDnonsmoker)
cm = st.CompareMeans(dsw1, dsw2)

# calculate number of observations, mean and standard deviation for each group
n_smoker = dsw1.nobs
n_nonsmoker = dsw2.nobs
mean_smoker = dsw1.mean
mean_nonsmoker = dsw2.mean
sd_smoker = dsw1.std
sd_nonsmoker = dsw2.std
print(f'n_smoker = {n_smoker}')
print(f'mean_smoker = {mean_smoker}')
print(f'sd_smoker = {sd_smoker}')
print()
print(f'n_nonsmoker = {n_nonsmoker}')
print(f'mean_nonsmoker = {mean_nonsmoker}')
print(f'sd_nonsmoker = {sd_nonsmoker}')
print()

# conduct hypothesis test
ht = cm.ztest_ind(alternative = 'two-sided', usevar = 'unequal', value = 0)

# calculate and print the standard error, the Z-score, and p-value for the hypothesis test
se = cm.std_meandiff_separatevar
testZ = ht[0]
p_value = ht[1]
print(f'Standard error = {se}')
print(f'Test statistic: Z = {testZ}')
print(f'p-value = {p_value}')

# reject or accept null hypothesis
if(p_value) < 0.05:
print('reject null hypothesis')
else:
print('accept null hypothesis')

n_smoker = 126.0
mean_smoker = 6.828730158730159
sd_smoker = 1.380668106117173

n_nonsmoker = 873.0
mean_nonsmoker = 7.144272623138601
sd_nonsmoker = 1.5178105512705897

Standard error = 0.13376049190705977
Test statistic: Z = -2.3590109449334946
p-value = 0.018323715325166526
reject null hypothesis


Exercise 5

Construct a confidence interval for the difference between the weights of babies born to smoking and non-smoking mothers.

1. Calculate a 95% confidence interval for the average length of pregnancies (weeks) and interpret it in context. Note that since you're doing inference on a single population parameter, there is no explanatory variable, so you can omit the x variable from the function.