Your table names will be unfamiliar to most everyone else. Aliases exist to help readability & comprehension ...
SELECT
AVG(a.Score) AS ScoreAvg,
COUNT(a.Score),
d.ID As DivisionID,
o.ID AS OrganizationID,
a.EmployeeID
FROM answers AS a
JOIN attrib_division_rl AS d USING(AttributionID)
LEFT JOIN attrib_organization_rl AS o USING(AttributionID)
LEFT JOIN employees AS tblEmp USING(AttributionID,EmployeeID)
GROUP BY DivisionID, OrganizationID, EmployeeID;
... assuming by CNT() you mean COUNT() and simplifying the employee Join.
Group By a Left Joined column contradicts itself ... Group By Null isn't possible, so either that column ought not to be Grouped By, or the Join needs to be Inner.
And Group By across joins multiplies aggregate results as if the Sorcerer's Apprentice had done them; you really want to try to do the aggregation in one base or derived table and Join to other tables separately for associated info as necessary.
The query Groups By joined columns d.id and o.id, so make a derived table ...
SELECT
d.ID As DivisionID, o.ID AS OrganizationID,
a.EmployeeID, a.Score
FROM answers AS a
JOIN attrib_division_rl AS d USING(AttributionID)
JOIN attrib_organization_rl AS o USING(AttributionID)
JOIN employees AS e USING(AttributionID,EmployeerID)
... and aggregate on that ...
SELECT
DivisionID, OrganizationID, EmployeeID,
AVG(a.Score) AS ScoreAvg, COUNT(a.Score)
FROM (
SELECT
d.ID As DivisionID, o.ID AS OrganizationID,
a.EmployeeID, a.Score
FROM answers AS a
JOIN attrib_division_rl AS d USING(AttributionID)
JOIN attrib_organization_rl AS o USING(AttributionID)
JOIN employees AS e USING(AttributionID,EmployeerID)
) x
GROUP BY DivisionID, OrganizationID, EmployeeID;