Skip to main content

Table 2 Default standardized mean differences (SMD) assigned to unpublished studies with continuous outcomes

From: SAMURAI: Sensitivity analysis of a meta-analysis with unpublished but registered analytical investigations (software)

Outlook Depends on published studies? If the outcome event is desired (higher.is.better = T) If the outcome event is undesired (higher.is.better = F)
‘very positive’ No 0.8 −0.8
‘positive’ 0.3 −0.3
‘no effect’ 0 0
‘negative’ −0.3 0.3
‘very negative’ −0.8 0.8
‘very positive CL’ Yes, depends on CL of published studies UCL, pub,1-α LCL, pub,1-α
‘positive CL’ 0.5 SMD ^ pub + UC L pub , 1 α 0.5 SMD ^ pub + LC L pub , 1 α
‘current effect’ SMD ^ pub SMD ^ pub
‘negative CL’ 0.5 SMD ^ pub + LC L pub , 1 α 0.5 SMD ^ pub + UC L pub , 1 α
‘very negative CL’ LCL, pub,1-α UCL, pub,1-α
  1. Note: ‘Positive’ outlooks favor the experimental treatment arm and ‘negative’ outlooks favor the control treatment arm.
  2. SMD: Standardized mean difference.
  3. LCL: Lower confidence limit.
  4. UCL: Upper confidence limit. SMD ^ pub : The summary SMD across all published studies.
  5. (LCL pub,1-α , UCL pub,1-a ): The 1-α confidence interval of SMD ^ pub .