BookmarkSubscribeRSS Feed
rola13
Calcite | Level 5

Hi there,

I used this code to calculate Kappa for two observers and a whole lot of categories:

proc freq
data= A_R_IOR;
tables act_A*act_R 
c_prone_A*c_prone_R 
chin_A*chin_R 
closed_A*closed_R 
e_asy_left_A*e_asy_left_R 
e_asy_right_A*e_asy_right_R 
e_axial_A*e_axial_R 
e_back_A*e_back_R 
e_droop_A*e_droop_R 
e_mov_A*e_mov_R 
e_OOS_A*e_OOS_R 
e_pinned_A*e_pinned_R 
e_for_A*e_for_R 
eyes_OOS_A*eyes_OOS_R 
face_ahead_A*face_ahead_R 
face_in_A*face_in_R 
face_left_A*face_left_R 
face_neutral_A*face_neutral_R 
face_out_A*face_out_R 
face_right_A*face_right_R 
face_straight_A*face_straight_R 
flat_side_A*flat_side_R 
folded_in_A*folded_in_R 
forehead_A*forehead_R 
friendly_give_A*friendly_give_R 
friendly_rec_A*friendly_rec_R 
front_legs_OOS_A*front_legs_OOS_R 
front_legs_stretched_A*front_legs_stretched_R 
front_legs_tucked_A*front_legs_tucked_R 
groom_self_A*groom_self_R 
half_closed_A*half_closed_R 
head_contact_floor_A*head_contact_floor_R 
head_contact_obsticle_A*head_contact_obsticle_R 
head_contact_own_body_A*head_contact_own_body_R 
head_mov_LY_A*head_mov_LY_R 
head_mov_S_A*head_mov_S_R 
head_no_contact_A*head_no_contact_R 
high_pressure_A*high_pressure_R 
hostile_give_A*hostile_give_R 
hostile_rec_A*hostile_rec_R 
inact_A*inact_R 
lying_A*lying_R 
lying_on_left_A*lying_on_left_R 
lying_on_right_A*lying_on_right_R 
mov_frequently_S_A*mov_frequently_S_R 
mov_occasionally_S_A*mov_occasionally_S_R 
neck_above_horizontal_A*neck_above_horizontal_R 
neck_below_horizontal_A*neck_below_horizontal_R 
neck_down_A*neck_down_R 
neck_horizontal_A*neck_horizontal_R 
neck_middle_A*neck_middle_R 
neck_mov_A*neck_mov_R 
nose_angle_normal_A*nose_angle_normal_R 
nose_angle_OOS_A*nose_angle_OOS_R 
nose_angle_stretched_A*nose_angle_stretched_R 
nose_A*nose_R 
nose_pressing_A*nose_pressing_R 
nothing_special_A*nothing_special_R 
one__front_leg_stretched_A*one__front_leg_stretched_R 
open_A*open_R 
rear_leg_stretched_A*rear_leg_stretched_R 
rear_legs_OOS_A*rear_legs_OOS_R 
rear_legs_tucked_A*rear_legs_tucked_R 
 rum_no_A*rum_no_R 
rum_OOS_A*rum_OOS_R 
rum_yes_A*rum_yes_R 
side_of_head_A*side_of_head_R 
slight_touch_A*slight_touch_R 
stand_fully_in_box_A*stand_fully_in_box_R 
stand_half_in_box_A*stand_half_in_box_R 
stand_in_alley_A*stand_in_alley_R 
stereo_A*stereo_R 
tail_directed_wagging_A*tail_directed_wagging_R 
tail_hanging_A*tail_hanging_R 
tail_lifted_straight_A*tail_lifted_straight_R
tail_not_lifted_A*tail_not_lifted_R 
tail_OOS_A*tail_OOS_R 
tail_OOS_lifted_A*tail_OOS_lifted_R 
tail_small_wagging_A*tail_small_wagging_R 
tail_strong_wagging_A*tail_strong_wagging_R 
tail_tucked_between_legs_A*tail_tucked_between_legs_R 
toung_rolling_A*toung_rolling_R/norow nocol nopercent;
test kappa;
run;

 

How do I convert what I see in the result viewer (sashtml) to an excel file? I only need the Kappa value!

 

Thank you 🙂

4 REPLIES 4
ballardw
Super User

First thing, the code you show will not generate ANY kappa values. You need to use the AGREE option on the tables statement to get them.

 

Second which "kappa value"?

The procedure would generate two tables, one for the Test statistic and the other Kappa coefficient.

 

Here is the approach for learning how to select any specific table(s) for output .

First run your code with the option ODS TRACE ON/off such as

ods trace on;

proc freq data=data;
   tables var1 * var2 / agree;
   test kappa;
run;

ods trace off;

The log will show some details about which tables were generated such as:

Output Added:
-------------
Name:       CrossTabFreqs
Label:      Cross-Tabular Freq Table
Template:   Base.Freq.CrossTabFreqs
Path:       Freq.Table1.CrossTabFreqs
-------------

Output Added:
-------------
Name:       McNemarsTest
Label:      McNemar's Test
Template:   Base.Freq.StatFactoid
Path:       Freq.Table1.McNemarsTest
-------------

Output Added:
-------------
Name:       Kappa
Label:      Simple Kappa Coefficient
Template:   Base.Freq.StatFactoid
Path:       Freq.Table1.Kappa
-------------

Output Added:
-------------
Name:       KappaTest
Label:      Simple Kappa Test
Template:   Base.Freq.StatFactoid
Path:       Freq.Table1.KappaTest
-------------

The key part you are looking for is the NAME.

 

You can then generate output for only the items you want by using ODS select such as:

ods select kappa;
proc freq data=data;
   tables var1 * var2 / agree;
   test kappa;
run;

will only generate the kappa table results. If want both the Kappa and the Kappatest use both names on the select.

To send to a different destination then you wrap the code generating the bits you want in an ODS destination "sandwich:

 

ods excel file="<system path>\file.xlsx"  <many options may go here>

<code that generates the output you want goes here
can be multiple procedures
>

ods excel close;

That CLOSE is what terminates sending the results to the ods destination and if not present you will not get the desired output.

 

You can specify different or multiple ODS destinations in a single program

 

rola13
Calcite | Level 5
Hi ballardw,
thank you so much for your detailed response. This is exactly what I needed! I checked again, if my code without the AGREE statement also gives me the simple kappa coefficient and it goes! Why did you think I wouldn't?
ballardw
Super User

@rola13 wrote:
Hi ballardw,
thank you so much for your detailed response. This is exactly what I needed! I checked again, if my code without the AGREE statement also gives me the simple kappa coefficient and it goes! Why did you think I wouldn't?

Slightly different versions of SAS most likely. Mine would not generate the output without agree.

Watts
SAS Employee

Hi rola13,

 

In your code, the KAPPA option in the TEST statement automatically invokes the AGREE option in the TABLES statement. (When you use only one TABLES statement, TEST statement options invoke the corresponding TABLES statement options -- documentation is here.)

 

The AGREE option in the TABLES statement provides kappa coefficients, standard errors and confidence limits. The KAPPA option in the TEST statement provides a kappa test (in addition to the kappa coefficients, standard errors, and confidence limits).

 

 

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

Mastering the WHERE Clause in PROC SQL

SAS' Charu Shankar shares her PROC SQL expertise by showing you how to master the WHERE clause using real winter weather data.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 4 replies
  • 609 views
  • 1 like
  • 3 in conversation