Re: Output format for p-values

[Douglas Bonett]

There is a bug in the computation of the p-value for the unequal variance option in Contrasts of One-way.  The output is shown below for a 3-group design with n = 3 per group and scores of 1,2,3,4,5,6,7,8,9.  The reported p-value is 1.98. 

ONEWAY /VARIABLES= score BY group

       /CONTRAST= 1 -1 0.

 

ANOVA

#====================#==============#==#===========#=====#====#

#                    #Sum of Squares|df|Mean Square|  F  |Sig.#

#=====#==============#==============#==#===========#=====#====#

#score|Between Groups#         54.00| 2|      27.00|27.00| .00#

#     |Within Groups #          6.00| 6|       1.00|     |    #

#     |Total         #         60.00| 8|           |     |    #

#=====#==============#==============#==#===========#=====#====#

 

Contrast Coefficients

#==========#==========================#

#          #           group          #

#          #--------+--------+--------#

#          #    1.00|    2.00|    3.00#

#========#=#========#========#========#

#Contrast|1#       1|      -1|       0#

#========#=#========#========#========#

 

Contrast Tests

#===============================#=============#========#=====#====#===========#

#                       Contrast#   Value of  |  Std.  |  t  | df |  Sig. (2- #

#                               #   Contrast  |  Error |     |    |  tailed)  #

#=====#================#========#=============#========#=====#====#===========#

#score|Assume equal    |    1   #        -3.00|     .82| 3.67|   6|        .01#

#     |variances       |        #             |        |     |    |           #

#     |Does not assume |    1   #        -3.00|     .82|-3.67|4.00|       1.98#

#     |equal           |        #             |        |     |    |           #

#=====#================#========#=============#========#=====#====#===========#

On Mon, Apr 21, 2014 at 10:33 AM, Douglas Bonett <address@hidden> wrote:

I am using PSPP in an undergrad social science stat course with 110 students. These students only use PSPP in GUI mode. We are evaluating PSPP as a possible replacement for SPSS in our undergraduate courses.  After two weeks of use, here are two issues that the students are complaining about:

1)  p-values are displayed in default F8.2 format and they need to report them with 3 decimal places. This will be another problem when we get to correlations and they need to report correlation estimates with 3 decimals (remember they are only using point and click options). Maybe changing to F7.3 for p-values and correlations would be an easy fix.

2)  when entering data, they get the following error message when then move to a new column:

                 warning: Data is not valid as format F: Number followed by garbage.

Both issues are minor but in the minds of these types of students they think their program is "broken" or "full of bugs".

Doug