How to Analyze Survey Data in Python: Simple Steps and Example

To analyze survey data in Python, use pandas to load and clean the data, then apply simple statistics like mean, median, and frequency counts with pandas and numpy. Visualization libraries like matplotlib or seaborn help to understand patterns visually.

Syntax

Use pandas.read_csv() to load survey data from a CSV file. Use DataFrame.describe() for summary statistics. Use value_counts() to count responses. Use matplotlib.pyplot or seaborn for charts.

• pd.read_csv('file.csv'): Load data
• df.describe(): Get stats like mean, std, min, max
• df['column'].value_counts(): Count unique answers
• plt.bar(), sns.countplot(): Visualize data

python

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Load survey data
survey_df = pd.read_csv('survey.csv')

# Summary statistics
summary = survey_df.describe()

# Count answers in a column
counts = survey_df['Question1'].value_counts()

# Plot counts
sns.countplot(x='Question1', data=survey_df)
plt.show()

Example

This example shows how to load survey data, get basic statistics, count answers for a question, and plot the results.

python

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Sample survey data as dictionary
data = {
    'Age': [25, 30, 22, 40, 28],
    'Satisfaction': ['Good', 'Excellent', 'Good', 'Poor', 'Excellent'],
    'Recommend': ['Yes', 'Yes', 'No', 'No', 'Yes']
}

# Create DataFrame
survey_df = pd.DataFrame(data)

# Show summary statistics for numeric columns
print(survey_df.describe())

# Count how many answered each satisfaction level
print(survey_df['Satisfaction'].value_counts())

# Plot satisfaction counts
sns.countplot(x='Satisfaction', data=survey_df)
plt.title('Survey Satisfaction Counts')
plt.show()

Output

             Age
count   5.000000
mean   29.000000
std     6.782330
min    22.000000
25%    25.000000
50%    28.000000
75%    30.000000
max    40.000000

Excellent    2
Good         2
Poor         1
Name: Satisfaction, dtype: int64

Common Pitfalls

Common mistakes include not cleaning data before analysis, such as missing values or inconsistent answers. Forgetting to convert data types (e.g., strings to numbers) can cause errors. Also, plotting without checking data can lead to confusing charts.

Always check for missing data with df.isnull().sum() and clean or fill missing values before analysis.

python

import pandas as pd

# Wrong: Not handling missing data
survey_df = pd.DataFrame({'Age': [25, None, 22], 'Satisfaction': ['Good', 'Excellent', None]})
print(survey_df.describe())  # May ignore missing data silently

# Right: Fill missing data before analysis
survey_df['Age'] = survey_df['Age'].fillna(survey_df['Age'].mean())
survey_df['Satisfaction'] = survey_df['Satisfaction'].fillna('Unknown')
print(survey_df.describe())

Output

count    2.000000
mean    23.500000
std      2.121320
min     22.000000
25%     22.750000
50%     23.500000
75%     24.250000
max     25.000000
dtype: float64
count          3
unique         3
top       Good
freq           1
dtype: object

count    3.000000
mean    23.666667
std      1.527525
min     22.000000
25%     22.833333
50%     23.666667
75%     24.500000
max     25.000000
dtype: float64
count          3
unique         3
top       Good
freq           1
dtype: object

Quick Reference

• Load data: pd.read_csv('file.csv')
• Summary stats: df.describe()
• Count values: df['col'].value_counts()
• Check missing: df.isnull().sum()
• Fill missing: df.fillna(value)
• Plot counts: sns.countplot(x='col', data=df)

Key Takeaways

Use pandas to load and summarize survey data easily.

Check and handle missing or inconsistent data before analysis.

Use value_counts() to count survey responses quickly.

Visualize data with seaborn or matplotlib for better insights.

Always verify data types and clean data for accurate results.