behaviour graph modifications

[srearl]

• [x] fix x axis to show all behaviours
• [x] fix y axis to 60%
• [x] order and color-code behaviours according to behaviour type (e.g., positive, negative); Anne will provide the categories

[srearl]

behaviour type corresponds to survey q3 but note that types are also present, mistakenly, in the database under q4 and needs review

[srearl]

the following four 2015 surveys are those that log behaviour type (e.g., Antisocial, Cognitive) in Q4 and should be omitted from a future implementation of the database and/or the error needs to be addressed, likely the latter as this is in keeping generally with problems associated with the 2015 surveys:

id,name,semester,year,class SV_0f70UyRiciL696Z,CSL-T TH Younger Class Data Observation Sheet,summer,2015,TTh_younger SV_3WLTQMpr41ekIRL,CSL-MWF Older Class Data Observation Sheet,summer,2015,MWF_older SV_41oVqbsApbpDzr7,CSL- MWF Younger Class Data Observation Sheet,summer,2015,MWF_younger SV_6gSMgf82lUhs1MN,CSL T Th Threes Data Observation Sheet ,spring,2015,TTh_three

surveys_with_behaviour_type_in_q4.csv

[srearl]

The aforementioned may not be correct. The problem data are 2015 but the following seems to be the problem, not the summer classes listed above. The above were generated from the initial approach of using modified date to identify year, class, semester, and the list in this comment is based on the newer approach that uses the metadata in the name field of each survey. Need to confirm.

response name count percent Antisocial Reliability Fall 2015 CSL Observation Sheets Pre-K - Copy 23 0.06 Antisocial Reliability Fall 2015 CSL Observation Sheets MWF Three's - Copy 6 0.01 Cognitive Reliability Fall 2015 CSL Observation Sheets Pre-K - Copy 11 0.03 Emotional Reliability Fall 2015 CSL Observation Sheets Pre-K - Copy 2 0.00 Emotional Reliability Fall 2015 CSL Observation Sheets MWF Three's - Copy 1 0.00 Prosocial Reliability Fall 2015 CSL Observation Sheets Pre-K - Copy 25 0.06 Prosocial Reliability Fall 2015 CSL Observation Sheets MWF Three's - Copy 4 0.01

...note that all of these have copy in the name field!

q4s_with_types.csv

[srearl]

UPDATE: both sets of surveys in the aforementioned comments have behaviour type woven into q4. The reason that the summer 2015 surveys were not caught in analyses identified by the search that revealed those from 2015 with copy in the field name is that the 2015 summer surveys do not have summer in the title, and the new filter only picks up surveys where spring, summer, or fall is in the title.

[srearl]

...which brings up the potential problem of surveys that do not have spring, summer, or fall in the name, such as spring 2015.