A broader question discusses the effect on MLU of any responses to questions. To try to answer this, let me share some stats on the % of C&IV (complete, intelligible, verbal) utterances that were responses to questions by grade taken from the SALT Conversation database:
Grade - % utts (resp to ques) - % utts (Y/N resp to ques)
PK - 46% - 21%
P - 37% - 13%
1 - 31% - 13%
2 - 31% - 12%
3 - 23% - 10%
5 - 24% - 11%
7 - 25% - 11%
If your sample contains significantly more or less responses to questions, you should consider excluding the responses to questions when comparing your sample to samples taken from the SALT Conversation database. When comparing MLU to developmental data taken from the literature, however, you need to know whether responses to questions were included.
A broader question discusses the effect on MLU of any responses to questions. To try to answer this, let me share some stats on the % of C&IV (complete, intelligible, verbal) utterances that were responses to questions by grade taken from the SALT Conversation database: Grade - % utts (resp to ques) - % utts (Y/N resp to ques) PK - 46% - 21% P - 37% - 13% 1 - 31% - 13% 2 - 31% - 12% 3 - 23% - 10% 5 - 24% - 11% 7 - 25% - 11% If your sample contains significantly more or less responses to questions, you should consider excluding the responses to questions when comparing your sample to samples taken from the SALT Conversation database. When comparing MLU to developmental data taken from the literature, however, you need to know whether responses to questions were included.