utility value of management tools
play

Utility value of management tools Advanced Herd Management Anders - PowerPoint PPT Presentation

Utility value of management tools Advanced Herd Management Anders Ringgaard Kristensen Slide 1 What kind of tools? Complex systems Not just single pieces of information Management information systems Bedriftslsningen


  1. Utility value of management tools Advanced Herd Management Anders Ringgaard Kristensen Slide 1

  2. What kind of tools? Complex systems • Not just single ”pieces of information” Management information systems • ”Bedriftsløsningen” • ”E-kontrol” Monitoring tools • Farm Watch Decision support tools Slide 2

  3. Why do we want to evaluate? Farmers who consider to buy a system would like to know the expected benefit Developers who wish to sell a system would like to be able to demonstrate the benefit Only very little research has been done in this field Slide 3

  4. Basic problems Value of single ”pieces of information” is difficult to assess. Secondary effects • Positive: Increased focus • Negative: Decreased focus in other areas The farmer perhaps doesn’t use the system in an optimal way. Interactions production system/farmer/tool No control (what would have happened without the tool?) Slide 4

  5. Methods (Verstegen et al. 1995) Slide 5

  6. Methods (Verstegen et al. 1995) Normative approaches • Decision theoretical approaches • Decision tree analysis • Bayesian Information Economics • Control Theory • Decision analytical approaches • Simulation • Linear and dynamic programming Slide 6

  7. Methods (Verstegen et al. 1995) � Normative approaches � Decision theoretical approaches � Decision tree analysis Not value of tools � Bayesian Information Economics � Control Theory � Decision analytical approaches � Simulation � Linear and dynamic programming Slide 7

  8. Methods (Verstegen et al. 1995) � Normative approaches � Decision theoretical approaches � Decision tree analysis Not value of tools � Bayesian Information Economics � Control Theory � Decision analytical approaches � Simulation Examples � Linear and dynamic programming Slide 8

  9. Dynamic programming Policy 1 2 3a 3b Milk yield, kg/cow/year 7 082 6 896 7 350 6 991 Average week of replac. 25 28 21 25 Annual replacement % 50 35 59 38 Net ret., DKK/cow/year 9 236 9 150 9 544 9 319 ” , DKK/(1000 kg milk) 1 304 1 327 1 299 1 333 Number of cows 100.0 102.7 96.4 101.3 Kristensen & Thysen (1991) Slide 9

  10. Dynamic programming Validity • What would the farmer do without the tool? • Would he/she follow the recommendations? • Are the registrations correct? • External validity: • Model versus real world • The tool tests itself • Bias for optimal policy Slide 10

  11. Simulation (Markov chain) Jalvingh et al. (1992) Slide 11

  12. Simulation (Markov chain) � Validity � What would the farmer do without the tool? � Would he/she follow the recommendations? � Are the registrations correct? � External validity: � Model versus real world � The tool tests itself � Bias for optimal policy Slide 12

  13. Simulation (Monte Carlo) Jørgensen & Kristensen (1995) Slide 13

  14. Simulation (Monte Carlo) � Validity � What would the farmer do without the tool? � Would he/she follow the recommendations? � Are the registrations correct? � External validity: � Model versus real world � The tool does not tests itself � No bias for optimal policy � The preferred normative approach Slide 14

  15. Empirical (”positive”) approaches Verstegen et al. (1995): • Experimental designs • Field experiments • Experimental Economics • Quasi-experimental designs • Non-experimental designs Use of data from herds Slide 15

  16. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 16

  17. Designs No control: • Only farms using the tool are included in the study Nonequivalent control (quasi-experimental design): • A control group is included in the study afterwards • As equal as possible to the farms using the tool True control (experiment in the usual sense) • Farms are randomly divided into two groups • One group is told to use the tool • The other group is not allowed to use it Slide 17

  18. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 18

  19. PO: Posttest only Result Not serious! Time Slide 19

  20. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 20

  21. NPO/TPO: Posttest only Result Manipulation • Confounding between farmer type, production system and use of tool } Effect Time Slide 21

  22. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 22

  23. PP: Pretest and posttest Result Manipulation • Perhaps a general trend: All farms may have improved as those being investigated } Effect • Confounding between general development and effect of tool Time Slide 23

  24. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 24

  25. NPP: Pretest and posttest Result Correction for • General trend • Confounding with farmer type (partially, } Effect no randomization) Time Slide 25

  26. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 26

  27. TPP: Pretest and posttest Result Correction for • General trend • Confounding with farmer type } Effect (randomization) Time Slide 27

  28. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 28

  29. TS: Time series, no control Result Confounding with farmer type } Effect Development over time • Value in the beginning versus full value Time Slide 29

  30. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 30

  31. NTS: Time series, control Result Development over time Effect: b - a • Value in the beginning } versus full value b No confounding with farmer type a Time Slide 31

  32. Classification of designs Time series Pretest- Posttest (TS) posttest only (PO) (PP) True control (N) TTS TPP TPO Nonequivalent NTS NPP NPO Control (N) No control TS PP PO Verstegen et al. (1995) Slide 32

  33. TTS: Time series, true control Result Development over time • Value in the beginning } Effect versus full value No confounding with farmer type Time Slide 33

  34. Example: NTS Value of a management information system for sow herds • Response: Piglets/sow/year • Nonequivalent control • Time series Slide 34

Recommend


More recommend