Presentation is loading. Please wait.

Presentation is loading. Please wait.

Christine Siegwarth Meyer Vice President Philadelphia, PA October 31, 2006 Effective uses of survey methods to determine damages Law Seminars International.

Similar presentations


Presentation on theme: "Christine Siegwarth Meyer Vice President Philadelphia, PA October 31, 2006 Effective uses of survey methods to determine damages Law Seminars International."— Presentation transcript:

1 Christine Siegwarth Meyer Vice President Philadelphia, PA October 31, 2006 Effective uses of survey methods to determine damages Law Seminars International Calculating & Proving Patent Damages FOR COLOR PRINTING: Under “File, Print,” first choose color printer, then “Properties, TekColor” tab, choose “Office Color, sRGB Display.” FOR B&W PRINTING: Choose B&W printer, but do not highlight Grayscale nor Pure B&W option—choose “Color.” For best B&W contrast with color background, convert file to a PDF.

2 Use of Surveys in IP Cases

3 2 Existing Data may not be Usable  Often, firm has no data  Data may not answer the relevant questions –Lost Profits What would plaintiff’s sales have been in the “but for” world? Did the patented feature drive demand for the product? Which non-infringing alternatives would have been acceptable to consumers? –Reasonable Royalty What portion of the realizable profit should be credited to the invention as distinguished from any non-patented elements, significant features or improvements added by the infringer?

4 3 Surveys in other Types of Cases  False advertising –Would sales have been lower if the allegedly false advertising was removed (but nothing else was changed)?  Trademark infringement –How many additional sales did the defendant get as a result of alleged use of a trademark?  Merger analysis –Do the two parties make products that are close substitutes? Many types of cases could benefit from a well-designed survey.

5 4 What Can a Well-Designed Survey Do?  Show the causal relationship, or nexus, between the change in the attribute at issue and the parties’ sales and profits  Hold constant all other factors that might influence the parties’ marketplace performance

6 Logit Survey

7 6  Five Steps in a Logit Survey –Questionnaire Development –Screening –Choice Ranking –Validation –Statistical Analysis of Results

8 7 Questionnaire Development  Identify relevant customers and potential customers  Identify likely substitute products  Identify likely important characteristics  Limit the burden on the respondent as much as possible

9 8 Screening  What is the relevant population? –Often, the population is all consumers who have bought or would consider buying that type of good  Where to best find the population  Screening questions  Consider using quotas

10 9 Choice Ranking  Show the respondent 3 to 4 items –Physical items or cards with pictures –Key identifying characteristics, including price  Item by item, ask the respondent whether he would consider purchasing the item as shown  Ask the respondent to identify their first choice, among the items shown

11 10 Validation  Survey firm calls respondents to verify participation  Observation of the survey by the expert  Expert should check for inconsistent answers

12 11 Statistical Analysis of Results  Several econometric and statistical techniques –Well-known in the economics literature –Developed, in part, by Nobel laureate, Daniel McFadden  Isolate the incremental effect of one product attribute on sales quantity and/or price  Analyze degree of substitutability between two products  Identify product to which customers would turn, if alleged infringer were not in the market

13 12 Benefits of the Logit Survey  Well-designed logit survey presents respondents with all possible substitute products –Firms often focus on a narrow set of competitors, while consumers may view a wider set products as alternatives  Can evaluate a “but-for” world that may not ever have actually existed –Survey design can isolate one factor and evaluate the incremental effect of changing that one factor  Can reduce or eliminate response bias –Surveys that ask “what would you do if …” questions are often inherently biased

14 Admissibility of Survey Evidence

15 14 Issues in Survey Design  The Manual for Complex Litigation lists seven criteria for deciding whether a survey is trustworthy –The population was properly chosen and defined. –The sample chosen was representative of the population. –The data gathered were accurately reported. –The data were analyzed in accordance with accepted statistical principles. –The questions were clear and not leading. –The survey was conducted by qualified persons following proper interview procedures. –The process was conducted so as to assure objectivity. Manual for Complex Litigation, 3rd Ed. (Washington, DC: U.S. Government Printing Office, 995), 101-103.

16 15 Issues in Survey Design  U.S. v. Dentsply (2003) lays out some general guidelines for survey design: –Survey must identify relevant respondents –Questionnaires must be based on results of pre-tests –Instructions must be clear, simple –Questions cannot be “leading” or suggest an answer –Response rates should be at least 50% and unbiased –Survey experts must be involved in the survey at all stages –Proper statistical analysis should be conducted –Statistical results must be fully replicable

17 How NOT to Conduct a Survey

18 17 The first 5,000 pages of the questionnaire were kind of wordy.  The survey in Dentsply was criticized because “[t]he instructions given to the respondents for completing the conjoint exercise were far too lengthy, complicated and difficult to understand.” (II, J, 2b)

19 18 Our interviewers “cleaned-up” the data.  The Dentsply survey was criticized because the firm conducting the survey changed some of the survey results. “Guideline Research engaged in the widespread practice of editing scenario cards that did not total 100 points and using those edited cards as data.” (II, J, 3c)

20 19 Our statistical software package was written “a long time ago, in a galaxy far, far away.”  In Dentsply, “Dr. Rossi was unable to replicate the PRIDEM software used to analyze the data derived from the DOJ’s survey. Prof. Wind regards PRIDEM as proprietary software. Prof. Wind asserted that the only way to access PRIDEM is to hire Prof. Wind or one of his associates.” (II, J, 3d)

21 Contact Information Christine Siegwarth Meyer Vice President White Plains, NY 10606 914-448-4119 Christine.meyer@nera.com © Copyright 2006 National Economic Research Associates, Inc. All rights reserved.


Download ppt "Christine Siegwarth Meyer Vice President Philadelphia, PA October 31, 2006 Effective uses of survey methods to determine damages Law Seminars International."

Similar presentations


Ads by Google