Biofam Family Life Trajectories Dataset

The biofam dataset contains family life trajectories for 2,000 individuals in Switzerland. It describes how people moved through different family-life states between ages 15 and 30, such as living with parents, leaving home, getting married, having children, or becoming divorced.

This dataset is a classic teaching dataset in social sequence analysis. It was constructed by Müller, Studer, and Ritschard (2007) from the retrospective biographical survey carried out by the Swiss Household Panel (SHP) in 2002. The same dataset is also distributed with the R package TraMineR as biofam. Our version follows the TraMineR version and is provided here so that Sequenzo users can learn sequence analysis with a well-known benchmark dataset in Python.

What does this dataset describe?

The dataset follows family life courses from age 15 to age 30. Each person is observed for 16 yearly time points.

In simple terms, each row is one person, and the yearly columns describe that person’s family-life situation at each age.

For example, one person may live with their parents for several years, then leave home, get married, and later have children:

Parent → Parent → Left → Left+Marr → Left+Marr+Child

Another person may remain living with their parents for the whole observed period:

Parent → Parent → Parent → Parent → Parent

This makes the dataset useful for learning how sequence analysis can describe family trajectories, compare life-course patterns, and study how family transitions differ across birth cohorts, gender, nationality, language groups, religion, or social background.

Data origin

• Source survey: Swiss Household Panel (SHP), retrospective biographical survey conducted in 2002.

• Dataset construction: The biofam dataset was constructed by Müller, Studer, and Ritschard (2007).

• TraMineR version: The dataset is included in TraMineR as biofam. According to the TraMineR documentation, it contains 2,000 rows, 16 state variables, 1 ID variable, 7 covariates, and 2 weight variables.

State sequence structure

The main sequence part of the dataset consists of 16 yearly family-life state columns.

These columns represent ages:

15, 16, 17, ..., 30

Each column records the person’s family-life state at that age.

For example, the sequence columns:

15, 16, 17, 18, ..., 30

should be read as:

family state at age 15, family state at age 16, family state at age 17, ..., family state at age 30

Family life states

The family-life states are coded from 0 to 7. These states are built from combinations of five basic family events or statuses:

Living with parents, left home, married, having children, divorced

The resulting state alphabet is:

Numeric Code	State Label	Description
0	Parent	Living with parents
1	Left	Left parental home
2	Married	Married, not yet left parental home in the combined coding
3	Left+Marr	Left parental home and married
4	Child	Having children
5	Left+Child	Left parental home and having children
6	Left+Marr+Child	Left parental home, married, and having children
7	Divorced	Divorced

In sequence analysis, these eight states form the alphabet of the dataset. An alphabet simply means the set of possible states that can appear in a sequence.

For example, the sequence:

0, 0, 0, 1, 1, 3, 6, 6

can be read as:

Parent → Parent → Parent → Left → Left → Left+Marr → Left+Marr+Child → Left+Marr+Child

This describes a gradual transition from living with parents, to leaving home, to marriage, and then to having children.

File structure

File: biofam.csv

The dataset contains:

Column type	Description
idhous	Individual or household identifier
Covariates	Background variables such as sex, birth year, nationality, language, religion, and parental social status
15 … 30	Yearly family-life states from age 15 to age 30
Weight variables	Two illustrative weight variables from the TraMineR version

A simplified structure looks like this:

idhous	sex	birthyr	nat_1_02	plingu02	15	16	17	...	30	wp00tbgp	wp00tbgs
66891	man	1943	Switzerland	german	0	0	0	...	6	1053.37	0.93
28621	man	1935	Switzerland	german	0	1	1	...	6	855.43	0.76
57711	woman	1946	Switzerland	french	0	0	0	...	6	575.21	0.51

Here, columns 15 to 30 are the actual sequence columns. They describe each person’s family-life state at each age.

Covariates

Besides the sequence columns, the dataset also includes background variables. These variables can be used to compare family trajectories across different social groups.

Column	Description
idhous	Individual or household identifier
sex	Sex of the respondent
birthyr	Birth year
nat_1_02	First nationality
plingu02	Language of the questionnaire
p02r01	Religion
p02r04	Religious participation
cspfaj	Father’s social status
cspmoj	Mother’s social status
wp00tbgp	Weight inflating to the Swiss population
wp00tbgs	Weight respecting sample size

Note on weights

The two weight variables are included for illustrative purposes only. Because biofam is a subsample of the original Swiss Household Panel data, these weights are not fully adapted to the actual dataset used here.

How to read the data

Suppose one row contains the following sequence from age 15 to age 30:

0, 0, 0, 0, 1, 1, 1, 3, 6, 6, 6, 6, 6, 6, 6, 6

Using the state table above, this means:

Parent → Parent → Parent → Parent → Left → Left → Left → Left+Marr → Left+Marr+Child → ...

Substantively, this person lived with their parents until early adulthood, then left home, then became married, and then entered the combined state of having left home, being married, and having children.

Another person may have the sequence:

0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0

This means that the person remained in the Parent state from age 15 to age 30.

These examples show why sequence analysis is useful. Instead of looking only at whether someone eventually married or had children, we can study the timing and order of multiple family-life transitions.

Why is this dataset useful?

The biofam dataset is useful because it is small, clear, and easy to interpret. It is especially suitable for beginners who want to learn the basic logic of state sequence analysis.

It can be used to learn how to:

• create a state sequence object;
• define a sequence alphabet;
• visualize individual family trajectories;
• plot state distributions by age;
• compare trajectories across birth cohorts or gender;
• compute sequence distances;
• cluster similar family-life courses;
• study how family transitions vary by social background.

For beginners, the key idea is that this dataset does not only record single events such as marriage or childbirth. It records a whole trajectory of family-life states from age 15 to 30. This allows us to study life courses as ordered sequences rather than isolated outcomes.

Typical research questions

With this dataset, we can ask questions such as:

• What are the most common family-life trajectories between ages 15 and 30?
• At what ages do people typically leave the parental home?
• Do men and women follow different family-life trajectories?
• Do family trajectories differ across birth cohorts?
• Are language, religion, nationality, or parental social status associated with different life-course patterns?
• Which trajectories involve early marriage or early parenthood?
• How common is the combined state of leaving home, marriage, and having children?

These questions are about the timing, sequencing, and duration of family-life states. This is exactly the kind of question that sequence analysis is designed to answer.

Relationship to TraMineR

In TraMineR, the dataset can be loaded as:

data(biofam)

The spell-format version can be loaded as:

data(bfspell)

The standard biofam data are in wide format: each row is one individual, and each age from 15 to 30 is represented by a separate column.

The bfspell version represents the same type of information in spell format, where each row corresponds to a spell or episode rather than a full individual sequence. For most introductory sequence-analysis tasks, the wide-format biofam dataset is easier to use. In Sequenzo, we currently only have wide-format version for simplicity; please do not hesitate to contact us if you would like us to incoporate more datasets in different formats on our GitHub Issue.

Reference

Müller, N. S., Studer, M., & Ritschard, G. (2007). Classification de parcours de vie à l'aide de l'optimal matching. In XIVe Rencontre de la Société francophone de classification (SFC 2007), Paris, 5–7 September 2007, pp. 157–160.

Swiss Household Panel: https://forscenter.ch/projects/swiss-household-panel/

TraMineR documentation: biofam, Example data set: Family life states from the Swiss Household Panel biographical survey.

Author: Yuqi Liang