LAST EMAIL

Single item questions - the stacked bar graphs look great, are you able to also me the % (+SD) of each response for the whole group and each group in a table please? - also if the change between BL and 12M is statistically significant for the whole group and for each treatment group?

No problem, I will provide you with measures of central tendency. However, for me this seems more like a categorical variable, so means, SD, medians and IQR may not be the most appropriate measures to summarise these variables. As for hypothesis testing, it is feasible to test differences between BL and 12M. Again, you should note that doing so for each treatment group may create some issues because of the low sample size, but it can be done. I will add those two analyses (for the whole sample and by group) in the next document. Depending of how we classify these variables (discrete or ordinal) we can use different tests.

Additionally, I attach the spreadsheet which has ’single item questions’. There were five questions asked at BL, 3M, 6M and 12M: 1. How is your health in general? (Single_GenHealth) 2. How is your dental health in general? (Single_DentHealth) 3. How is your periodontal health in general? (Single_PerioHealth) 4. Since completing periodontal treatment, has your periodontal health…? ChangeinPerioHealth_PostTx - n.b this only has 3M, 6M and 12M 5. To What extend has these problems affected your quality of life overall? (ExtentAffectQoL)

For these questions I’d like

by treatment group, how response has changed over time (i.e. within group) and comparing between groups. Is there a statistically significant difference between groups with regard to change between each time-point? Is this best depicted in a bar graph?
Proportions of each response in the whole sample at each time-point and by treatment group (bar graph)?

Loading data

Selecting required vars

Description

Summary measures

an overall picture (of the whole sample) of how their response has change over time (i.e. BL, 3M and 12M). Is the change statistically significant? A line graph plotting each participant ideally

Stacked bar:

General health

DenHealth

PerioHealht

ExtentAffectQoL_BL

Change in perio

Change over time

Table

Plotting line and bar

Change over time by treatment group

by treatment group, how response has changed over time (i.e. within group) and comparing between groups. Is there a statistically significant difference between groups with regard to change between each time-point? Is this best depicted in a bar graph?

General health by tx

Dental health by tx

Perio Health by tx