I would like your help with SQL to drive a report on customer survey responses. The main problem is needing to account for responses that were NOT given.

-- Dim table containing all potential survey responses
--
create table survey_response_dim
( survey_response_dim_pk number(10)
, survey_name char(8)
, question_code char(4)
, response_code char(4)
, response_descr varchar2(60));

-- Fact table containing customer survey responses
--
create table cust_survey_resp_fct
( customer_pk number(10)
, survey_response_dim_pk number(10)
, response_date date);

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 1, 'S1', 'Q1', 'F', 'FEMALE' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 2, 'S1', 'Q1', 'M', 'MALE' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 3, 'S1', 'Q2', 'A', '0-10' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 4, 'S1', 'Q2', 'B', '11-30' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 5, 'S2', 'Q1', '1', 'Strongly Agree' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 6, 'S2', 'Q1', '2', 'Agree' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 7, 'S2', 'Q1', '3', 'Disagree' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 8, 'S2', 'Q1', '4', 'Strongly Disagree' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 9, 'S1', 'Q3', 'Y', 'Yes' );

INSERT INTO survey_response_dim
( survey_response_dim_pk, survey_name, question_code, response_code, response_descr )
VALUES ( 10, 'S1', 'Q3', 'N', 'No' );

-- Fact table containing customer survey responses
--
create table cust_survey_resp_fct
( customer_pk number(10)
, survey_response_dim_pk number(10)
, response_date date);

INSERT INTO cust_survey_resp_fct
( customer_pk, survey_response_dim_pk, response_date )
VALUES ( 10, 2, to_date('05/21/2004', 'mm/dd/yyyy') );

INSERT INTO cust_survey_resp_fct
( customer_pk, survey_response_dim_pk, response_date )
VALUES ( 10, 4, to_date('05/21/2004', 'mm/dd/yyyy') );

INSERT INTO cust_survey_resp_fct
( customer_pk, survey_response_dim_pk, response_date )
VALUES ( 20, 3, to_date('05/21/2004', 'mm/dd/yyyy') );

INSERT INTO cust_survey_resp_fct
( customer_pk, survey_response_dim_pk, response_date )
VALUES ( 30, 2, to_date('05/21/2004', 'mm/dd/yyyy') );

INSERT INTO cust_survey_resp_fct
( customer_pk, survey_response_dim_pk, response_date )
VALUES ( 30, 3, to_date('05/21/2004', 'mm/dd/yyyy') );

INSERT INTO cust_survey_resp_fct
( customer_pk, survey_response_dim_pk, response_date )
VALUES ( 40, 6, to_date('05/22/2004', 'mm/dd/yyyy') );


Select sd.survey_name, sd.question_code
, sd.response_code, count(*) as responders
from cust_survey_resp_fct s
, survey_response_dim sd
where s.survey_response_dim_pk = sd.survey_response_dim_pk
group by sd.survey_name, sd.question_code, sd.response_code


SURVEY_NAME QUESTION_CODE RESPONSE_CODE RESPONDERS
S1 Q1 M 2.00
S1 Q2 A 2.00
S1 Q2 B 1.00
S2 Q1 2 1.00


I would like to also account for responses that were NOT given. The report would look similar to:

SURVEY_NAME QUESTION_CODE RESPONSE_CODE RESPONDERS
S1 Q1 F 0.00
S1 Q1 M 2.00
S1 Q2 A 2.00
S1 Q2 B 1.00
S1 Q3 N 0.00
S1 Q3 Y 0.00
S2 Q1 2 1.00

You addressed a similar issue (missing dates in a transaction table) in a recent article in Oracle Magazine but I just didn't "get" it. Hopefully, I'll clue in this time. Thanks for your help.