Problem
Given a list of feedbacks to questionary, output the statistics report, grouped by ethnic group.
One feedback record contains info like ethnic group, gender, citizenship, etc.

Solution
This is a simple problem. The key here is to choose and clearly define the data structures. Instinctively,

1. Feedback record: One feedback record is naturally a dictionary like {‘ethnic’:'African American’, ‘gender’:'Male’, ‘citizenship’:'US’ }.
2. Statistics: It is a list of statistics record, where each record is identified by ethnic group and contains several statistics fields. Likewise, one statistics record is naturally a dictionary like {‘ethnic’:'African American’, ‘Gender_Male’:3, ‘Gender_Female’:3, ‘citizenship_US’: 4, ‘citizenship_permvisa’: 1, …}

Side note: the statistics table can be generated by SQL-like statement (cause it’s actually aggregate query) like: select ethnic, count(gender == ‘male’) as gender_male, … group by ethnic

Code:
Based on data structure definition, code is straightforward:

def prog_3_A_questionary(l):
    """
    >>> l = []
    >>> l.append(dict(ethnic='African American', gender='Male', citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='African American', gender='Male', citizenship='US Citizen'))
    >>> l.append(dict(ethnic='African American', gender='Male', citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='African American', gender=None, citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='African American', gender='Female', citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='African American', gender='Female', citizenship='Temp'))
    >>> l.append(dict(ethnic='African American', gender='Female', citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='African American', gender='Female', citizenship=None))
    >>> l.append(dict(ethnic='Spanish Surname', gender='Female', citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='Spanish Surname', gender='Female', citizenship='Temp'))
    >>> l.append(dict(ethnic='Spanish Surname', gender='Female', citizenship='Perm Visa'))
    >>> l.append(dict(ethnic='Asian American', gender='Male', citizenship='Perm Visa'))
    >>> from pprint import pprint
    >>> pprint(prog_3_A_questionary(l))
    {'African American': {'Total': 8,
                          'citizenship_Perm Visa': 5,
                          'citizenship_Temp': 1,
                          'citizenship_US Citizen': 1,
                          'ethnic_African American': 8,
                          'gender_Female': 4,
                          'gender_Male': 3},
     'Asian American': {'Total': 1,
                        'citizenship_Perm Visa': 1,
                        'ethnic_Asian American': 1,
                        'gender_Male': 1},
     'Spanish Surname': {'Total': 3,
                         'citizenship_Perm Visa': 2,
                         'citizenship_Temp': 1,
                         'ethnic_Spanish Surname': 3,
                         'gender_Female': 3}}
    """
    # TODO: actually, we should write code to initialize the stat fields like
    # 'Gender_Male' to be 0. This will help:
    # - reduce the runtime check logic to see if the field is already there or not
    # - initialize all the stat_fields to 0, in order to prevent any missing stat_fields.
    from collections import defaultdict
    stat = {}
    for record in l:
        # get stat record by 'group key' - ethnic
        ethnic = record['ethnic']
        if not ethnic in stat:
            stat_record = {}
            stat[ethnic] = stat_record
        stat_record = stat[ethnic]
        # increase 'Total'
        if not 'Total' in stat_record:
            stat_record['Total'] = 0
        stat_record['Total'] += 1
        # increase stat fields like gender_Male
        for key, val in record.items():
            # handle missing answer
            if val == None:
                continue
            stat_field = '%s_%s' % (key, val)
            if not stat_field in stat_record:
                stat_record[stat_field] = 0
            stat_record[stat_field] += 1
    return stat