Campbell et al v. Facebook Inc.

Filing 166

Administrative Motion to File Under Seal re Plaintiffs' Reply in Support of Class Certification filed by Matthew Campbell, Michael Hurley. (Attachments: # 1 Proposed Order, # 2 Certificate/Proof of Service, # 3 Declaration of Melissa Gardner, # 4 Exhibit A (filed under seal), # 5 Exhibit B (redacted), # 6 Exhibit C (filed under seal), # 7 Exhibit D (redacted), # 8 Exhibit E (filed under seal), # 9 Exhibit F (redacted), # 10 Exhibit G (filed under seal), # 11 Exhibit H (redacted), # 12 Exhibit I (filed under seal), # 13 Exhibit J (redacted), # 14 Exhibit K (filed under seal), # 15 Exhibit L (filed under seal), # 16 Exhibit M (filed under seal), # 17 Exhibit N (filed under seal), # 18 Exhibit O (filed under seal), # 19 Exhibit P (filed under seal), # 20 Exhibit Q (filed under seal), # 21 Exhibit R (filed under seal), # 22 Exhibit S (filed under seal), # 23 Exhibit T (filed under seal), # 24 Exhibit U (filed under seal), # 25 Exhibit V (filed under seal), # 26 Exhibit W (filed under seal))(Gardner, Melissa) (Filed on 2/19/2016)

EXHIBIT D
REDACTED VERSION OF
DOCUMENT(S) SOUGHT TO BE SEALED

EXHIBIT 1

REDACTED VERSION OF DOCUMENT(S) SOUGHT TO BE SEALED
1
2
3
4
5
6
7

Michael W. Sobol (State Bar No. 194857)
[e-mail]
David T. Rudolph (State Bar No. 233457)
[e-mail]
Melissa Gardner (State Bar No. 289096)
[e-mail]
LIEFF CABRASER HEIMANN & BERNSTEIN, LLP
275 Battery Street, 29th Floor
San Francisco, CA 94111-3339
Telephone: [phone]
Facsimile: 415.956.1008

12

Hank Bates (State Bar No. 167688)
[e-mail]
Allen Carney
[e-mail]
David Slade
[e-mail]
CARNEY BATES & PULLIAM, PLLC
11311 Arcade Drive
Little Rock, AR 72212
Telephone: [phone]
Facsimile: 501.312.8505

13

Attorneys for Plaintiffs and the Proposed Class

8
9
10
11

14

UNITED STATES DISTRICT COURT

15

NORTHERN DISTRICT OF CALIFORNIA

16
17
18

MATTHEW CAMPBELL and MICHAEL
HURLEY, on behalf of themselves and all
others similarly situated,

19

Plaintiffs,

20

FACEBOOK, INC.,

REBUTTAL REPORT OF DR. JENNIFER
GOLBECK IN SUPPORT OF
PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION
_______________

v.

21

Case No. C 13-05996 PJH (MEJ)

22
23

Defendant.

HEARING
Date: March 16, 2016
Time: 9:00 a.m.
Place: Courtroom 3, 3rd Floor
|
The Honorable Phyllis J. Hamilton

24
25
26
27
28
REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

TABLE OF CONTENTS

2

Page

3
4
5
6

I.
II.
III.

7
8

IV.

9
10
11

V.
VI.

12

VII.

SCOPE OF ENGAGEMENT ................................................................................. 1
METHODOLOGY.................................................................................................. 1
ASCERTAINABILITY .......................................................................................... 2
A.
Class Members are Ascertainable from Facebook’s Records ..................... 2
FACEBOOK’S INTERCEPTION OF PRIVATE MESSAGE CONTENT........... 9
A.
Are not Necessary for Message Delivery .................................. 9
B.
.............................................................................. 11
ALLEGED “VARIABILITY” .............................................................................. 13
WHETHER PRIVATE MESSAGE CONTENT WAS INTERCEPTED IN
TRANSIT OR IN STORAGE............................................................................... 15
FACEBOOK’S USE OF PRIVATE MESSAGE CONTENT IN THE
SOCIAL GRAPH AND TARGETED ADVERTISING ...................................... 15

13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
-i-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

I.

SCOPE OF ENGAGEMENT
1.

2

I have been asked by the Plaintiffs through their counsel to respond to the

3

conclusions expressed in the Expert Report of Dr. Benjamin Goldberg submitted with Defendant

4

Facebook Inc.’s Opposition to Plaintiffs’ Motion for Class Certification (“Goldberg Report”),

5

statements made by Facebook Engineering Manager Alex Himel in his declaration submitted in

6

support of Facebook’s Opposition, and characterizations of my testimony made by Facebook in

7

its Opposition to Plaintiffs’ Motion for Class Certification.

8

II.

9

METHODOLOGY
2.

My rebuttal opinions, as well as the evidence I rely upon to support them, are set

10

forth in detail in this rebuttal report. The contents of the various exhibits that I identify by name

11

are meant to be incorporated, in their entirety, by such reference.

12

3.

As with my opening report submitted in connection with Plaintiffs’ Motion for

13

Class Certification (“Golbeck Opening Report”), in preparing this report, I have employed

14

methods and analyses of a type reasonably relied upon by experts in my field in forming opinions

15

or inferences on the subject. The opinions expressed are based upon a reasonable degree of

16

computer science certainty.

17

4.

Between now and such time that I may be asked to testify before the Court, I

18

expect to continue my review, evaluation, and analysis of information generated during

19

discovery, as well as of relevant evidence presented before and/or at trial. I also expect to review

20

any further reports submitted by Facebook’s experts. I reserve the right to amend or supplement

21

this rebuttal report, as necessary and as acceptable to the Court. I also reserve the right to develop

22

materials and exhibits as appropriate for use in helping to demonstrate and explain my opinions in

23

the event that I am asked to testify at trial.

24

5.

In forming my opinions, I have reviewed source code which I understand was

25

provided by Facebook’s counsel and which was represented as containing the relevant source

26

code between some time in 2009 and December 2012.

27
28

6.

Additionally I have reviewed internal Facebook documents produced in this

litigation, the Goldberg Report, the transcript of the deposition of Dr. Goldberg, the declarations
-1-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

of various Facebook employees submitted in in support of Facebook’s Opposition to Plaintiffs’

2

Motion for Class Certification, the deposition transcripts of those employees, as well as certain

3

public materials. The list of documents I have considered in forming my opinions in this rebuttal

4

report is attached to this report as Appendix A.

5

III.

ASCERTAINABILITY

6

A.

Class Members are Ascertainable from Facebook’s Records

7

7.

Facebook has a database called

8

. I understand that Facebook produced the

9

Exhibit A to

10

Facebook’s Second Supplemental Responses and Objections to Plaintiffs’ Narrowed Second Set

11

of Interrogatories

12
13
14
15

.

16

8.

The contents of the page include

9.

The

, including:

17
18
19
20
21
22

can be used to access information about Class members. The

23
24
25

could be written that would identify the senders and
recipients of Private Messages sent during the Class Period with URL attachments

26

by doing the following:

27
28

1
2

Facebook Appendix (“App.”) at 1534-1555.
Ex. 7 to the Declaration of David Slade (“Slade Decl.”)
-2-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1
2
3
4
5
6
7

10.

Specifically, the query would involve identifying the following information related

to the Class Definition:

8

a.

Private Messages within the Class Period based upon

9

(highlighted in yellow in Slade Decl. Ex. 7);

10

b.

Within that time period, Private Messages that

11
12
13

(highlighted in red in Slade Decl. Ex. 7);

14

c.

Sender of each Class-qualifying message based upon

15
16

(highlighted in blue in Slade Decl. Ex. 7); and

17

d.

Recipient(s) of each Class-qualifying message based upon

18
19

(highlighted in green in Slade
Decl. Ex. 7).

20
21

11.

Focusing on the above-described fields also addresses an issue raised by Facebook

in its brief: the fact that
.3 In discovery, I understand that Facebook was able to produce the

22
23
24
25

, marked
as FB000005577, FB000005800, FB000005882, FB000006007, FB000006088, FB000012006,

26
27
28

3

Opp. at 12:11.
-3-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

and FB000012557.4 Reviewing each of these

2
3

:

4

a.

FB0000055775 shows that

5
6
7

in the course of the transmission of this Private
Message, and the above-describe query I articulate would exclude such a message, accordingly.

8

b.

FB0000058006 presents a similar case in that,

9
10
11

this
message would be between Facebook users who were outside of the proposed Class.

12

c.

FB000005882,7 like FB000005577, shows that

13
14

.

15

d.

FB000006007,8 like FB000005577, shows that

16
17

.

18

e.

FB000006088,9 like FB000005577, shows that

19
20

.

21

f.

FB000012006,10 like FB000005577, shows that

22
23
24
25
26
27
28

.
4

This understanding is based upon the table represented in Exhibit A To Defendant Facebook,
Inc.’s Second Supplemental Responses And Objections To Plaintiffs’ Narrowed Second Set Of
Interrogatories (App. 1534-1555).
5
Slade Decl. Ex. 14.
6
Slade Decl. Ex. 15.
7
Slade Decl. Ex. 16.
8
Slade Decl. Ex. 17.
9
Slade Decl. Ex. 18.
10
Slade Decl. Ex. 19.
-4-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

g.

FB00001255711 displays a

2
3
4

, as this message would be between Facebook users who
were outside of the proposed Class.

5

12.

The above query identifies the relevant fields within

6

. As I describe in paragraphs 98-105 of my Opening Report,

7
8
9

, Class members can be readily identified.

10

The above query addresses what I understand to be the relevant inquiry for identifying Class

11

members: that is, whether or not

12

URL attachment.

13

will separate Class members from non-Class members .

14

13.

If the names of

15
16

have subsequently changed, this query could be modified
accordingly to address any changes

17
18

a Private Message sent with a

14.

.

It appears that Dr. Goldberg’s and Facebook’s criticism of my methods described

in my opening report and my deposition are based on an assumption that

19
20

. Although I did not

21
22

to make these identifications, even offering an example query
(though it may need some tweaking after I have the opportunity to test it in practice).

23

15.

As I state in my Opening Report, the queries I offered were not intended to

24

retrieve a final list of Class members. As a preface to the sample code, I stated, “[a] database

25

query could be used

26

(¶ 103). I never suggested that everyone

27
28

was equivalent to the complete list of the Class members.
11

Slade Decl. Ex. 20.
-5-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

16.

Dr. Goldberg and Mr. Himel both argue my example code will be both under- and

2

over- inclusive of Class members, without acknowledging that the parameters necessary to

3

identify the Class are readily available.

4

17.

At ¶¶ 66-77 of his report Dr. Goldberg argues that the method that I propose would

5

(id. ¶

6

67). However, the examples that he provides are cases which either 1) take the user out of Class

7

definition or, 2) are due to system failures, the frequency of which is likely very low. I respond to

8

each of these cases below:

9
10

a.

¶ 68: “This query will be under-inclusive in that it will not reflect recipients

of messages

11
12

.
b.

¶ 69: “This query will be under-inclusive in that it will not identify

13
14

I have not seen any
evidence or documentation supporting Dr. Goldberg’s underlying assumption that if

15
16
17
18

c.

¶ 70: “This query will be under-inclusive in that it will not identify senders

19

and recipients whose accounts were deleted.” I have not seen any evidence or documentation

20

supporting Dr. Goldberg’s underlying assumption that

21
22
23
24

.
d.

¶ 71: “This query will be under-inclusive in that it will not identify

25
26
27
28
-6-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1
2

.

3

e.

¶ 72: “This query will be under-inclusive in that it will not identify

f.

¶ 73: “This query will be under-inclusive in that it will not identify

4
5
6
7

.

8
9
10
11
12

.

13

g.

¶ 74: “This query will be over-inclusive in that it will include senders

14

whose messages did not contain URLs in their text.” This appears to be referencing a scenario

15

where

16
17
18
19
20

.
h.

¶ 75: “This query will be over-inclusive in that it will include

21
22
23
24
25
26

.
i.

¶ 76: “This query will be over-inclusive in that it will include senders and

27

recipients outside the United States.” Facebook knows if users are within the United States and

28

could check this for senders and recipients whose messages were intercepted. Facebook states on
-7-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

its website that this data is collected for advertising purposes.12

2
3
4
5
6

.

7
8

j.

¶ 77: “This query will be over-inclusive in that it will include senders of

messages outside the Class Period.” As explained above,

9

.

10

18.

At ¶¶ 78-86 of his report, Dr. Goldberg argues that my proposed methods are

11

“overbroad in that [they] will identify senders that were not subject to the challenged ‘uses.’”

12

However, the Facebook code is written such that

13
14
15
16

.

17

19.

At ¶ 78 of his report, Dr. Goldberg states: “Dr. Golbeck’s query is overbroad in

18

that it will identify senders that were not subject to the challenged ‘uses.’ In her deposition, Dr.

19

Golbeck conceded each of these flaws in her proposed query and said that identifying those that

20

were subject to the challenged ‘uses’ would be ‘case-specific.’” This is a misstatement of my

21

deposition testimony. I was asked specifically if

22
23

not if it was possible to
identify senders subject to uses.14

24
25
26
27
28

12

See https://www.facebook.com/business/help/133609753380850 (“How does Facebook know
when people are in the locations I’m targeting? Facebook uses information from multiple
sources such as current city from profile, IP address, data from mobile devices if location services
are enabled, and aggregated information about the location of friends.”).
13
FB000027191.
14
App. 1337 (Golbeck Depo. 344:7-19).
-8-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

20.

At ¶ 79 of his report, Dr. Goldberg states: “This query cannot identify senders

2
3
4

.

5
6

21.

Similarly, the “uses” Dr. Goldberg identifies at ¶¶ 80-8615 of his report are all

instances where

7
8
9
10
11

.
IV.

12

FACEBOOK’S INTERCEPTION OF PRIVATE MESSAGE CONTENT
A.

13

Are not Necessary for Message Delivery
1.

14
22.

Dr. Goldberg argues that creating objects in object-oriented programming

15
languages is common practice. This is true. However, objects are not side effects of object16
oriented programming. They are data structures that people create and that are explicitly
17
programmed to record information. Thus, if data is stored in an object, it is because a programmer
18
made an explicit decision to record that information. If a system intercepts content from
19
communications, it is not part of the ordinary course of business simply because the intercepted
20
content is stored in an object.
21
23.

Dr. Goldberg argues that

22
23
24
25

15

26
27
28
.
-9-

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

Information stored in a

2

database is not part of the code. While the stored data may map to what is implemented in objects

3

in the code, it is not an object-oriented programming object itself.

4

2.

5

24.

Alternative methods

There are alternative methods for

6

.

7

a.

For example, rather than

8
9
10

.

11

b.

12

Additionally,
.

13

c.

The fact that these alternative methods of

14
15
16
17
18
19
20
.17

21
22

3.

23

25.

Code-Based Devices

Dr. Goldberg argues that he has never heard the term "code-based device" before

24

(¶ 8). However, code-based or software devices are quite common, and just because Dr. Goldberg

25

has not heard the term does not mean they cannot exist.

26
27
28

16

See, e.g. section “Dissection of an Object” in Hasin Hayder, Object-Oriented Programming
with PHP5(Packt Publishing Ltd, 2007) .
17
See App. 1522-23 (Himel Decl. ¶ 44); App. 1697-98, 1699-1700, 1702 (Fechete Decl. ¶¶ 1314, 18, 26).
- 10 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

26.

For example, one domain where the public is hearing a lot about code-based

2

devices now is in the Volkswagen emissions controversy.18 Indeed, a Google News search for

3

“Volkswagen 'software device'” returned about 37,400 results.19 I understand that the

4

Environmental Protection Agency has stated that software can constitute a “device” as that term

5

is used in government regulations. 20

6
7

27.

The term "software device" has also appeared in US patents21 and publications

from NASA.22

8

B.

9

28.

I understand that Mr. Himel claims that

29.

I analyzed

10
11
12
13
14

.

15

30.

Based upon

16

.

17

31.

In

18
19
20
21
22
23
24
25
26
27
28

:
18

See Slade Decl. Ex 12 (Goldberg Depo. Tr. at 171:10-177:22.)
Slade Decl. Ex 13.
20
See id.; see also “EPA, California Notify Volkswagen of Clean Air Act Violations / Carmaker
allegedly used software that circumvents emissions testing for certain air pollutants,” available at
http://yosemite.epa.gov/opa/admpress.nsf/a883dc3da7094f97852572a00065d7d8/dfc8e33b5ab16
2b985257ec40057813b!OpenDocument (“As described in the [Notice of Violation], a
sophisticated software algorithm on certain Volkswagen vehicles detects when the car is
undergoing official emissions testing, and turns full emissions controls on only during the test . . .
The software produced by Volkswagen is a “defeat device,” as defined by the Clean Air Act.”)
21
See, e.g. , U.S Patent No. 6,032,223.
22
See, e.g., V. Kreinovich, A. Bernat, E. Villa, Y. Mariscal, “Parallel computers estimate errors
caused by imprecise data,” Interval Computations, 1991, No. 2, pp. 31–46. (available at
http://ntrs.nasa.gov/search.jsp?R=19930068753).
23
See App. 1522-23 (Himel Decl. ¶ 44).
24
See, e.g., FB000014213; FB000027011; FB000027015; FB000027018.
19

- 11 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1
2
3
4

32.

5
6
7
8
9
10
11
12

33.

These show

13

34.

Even if this information is not

35.

Dr. Goldberg stated in his deposition that he

.

14
15
16
17
18
19
20
21
22
27

23
24

36.

I note that, in the context of Facebook’s

25
26
27
28

25

.

26

Slade Decl. Ex. 12 (Goldberg Depo. Tr. at 139:10-143:6).
27
FB000027190.
- 12 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1
2
3
4
5

37.

Indeed, the only

6
29

7
8

V.

9

ALLEGED “VARIABILITY”
38.

At paragraphs 60 and 63 of his report, Dr. Goldberg argues that alleged

10

“variability” in what he calls “interceptions” and “uses” of Private Message content would require

11

a “a message-by-message analysis” to determine “whether such alleged interceptions” or “uses”

12

occurred. Mr. Himel addresses these same “variabilities” in his declaration.30 However, most of

13

these “variabilities” simply track the same scenarios that Dr. Goldberg incorrectly argues makes

14

identification of Class members impossible. As I explain in Section III of the this rebuttal report,

15
16

.

17

39.

I note that while Dr. Goldberg claims that

18
19
20
21
.32

22
23
24
25
26
27
28

28

App. 1943 (Goldberg Report, ¶ 9).
See App. 1964 (Goldberg Report ¶ 44); App. 1697-98, 1699-70, 1702 (Fechete Decl. ¶¶ 13-14,
18, 26); Golbeck Opening Report ¶¶ 44-54.
30
See, generally App. 1508-33 (Himel Decl.)
31
Slade Decl. Ex. 12 (Goldberg Depo. Tr. at 80:21-23)
.
Slade Decl. Ex. 12 (Goldberg Depo. Tr. at 81:8-86:17).
29

- 13 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

40.

Several of these additional variabilities do not overlap with the scenarios Dr.

2

Goldberg incorrectly argues makes identification of Class members impossible. I address each as

3

follows:

4

a.

5
6
7
8
9

and therefore the message would not be within the Class
definition.
b.

10
11
12
13

and therefore the message would not be within the Class definition.
c.

14
15
16

and therefore the message would not be within the

17

Class definition.

18

d.

19
20
21
22

and
are not relevant to this action.
e.

In summary,

23

. Also, Dr. Goldberg and Mr.

24

Himel are correct that not all Private Messages, or even all Private Messages that include URLs,

25

are included within the Class. However, these “variabilities” do not cause any complexities

26

related to identifying Class members or resolving issues related to Class members because each

27

of these “variabilities”

28

.
- 14 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

VI.

2

WHETHER PRIVATE MESSAGE CONTENT WAS INTERCEPTED IN
TRANSIT OR IN STORAGE
41.

Messages are in transit when they are intercepted. They are delivered

3
4
5

.33 Dr.
Goldberg argues that messages that are in memory are in “storage” and that he has never heard of
the two being distinguished in any context. However, the distinction is commonly made.

6
7
8
9

42.

does PC Magazine35 and numerous other websites.36 Page 370 of the textbook, “Discovering
Computers,” by Misty E. Vermaat, et al. distinguishes them.37 While Dr. Goldberg may be
unaware that these are treated differently, such a distinction does in fact exist.

10
11
12

As a few examples, Microsoft distinguishes the two on their Windows website34 as

43.

Indeed, if Dr. Goldberg's position were correct, it would be impossible for a

computer-based violation of wiretap law, since computers must have data in memory in order to
operate on it.

13
14

VII.

FACEBOOK’S USE OF PRIVATE MESSAGE CONTENT IN THE SOCIAL
GRAPH AND TARGETED ADVERTISING

15

44.

I note that Facebook states the following concerning my deposition testimony:

16
17
.”38

18
19
20
21
22
23
24
25
26
27
28

33

See Golbeck Opening Report ¶¶ 30-31.
See “Memory and storage,” available at http://windows.microsoft.com/enus/windows7/memory-and-storage.
35
See “Definition of: storage vs. memory,”
http://www.pcmag.com/encyclopedia/term/63352/storage-vs-memory (“The difference between
storage and memory is that non-volatile storage is used to hold programs and data until purposely
changed or removed by the user, while volatile memory is a temporary workspace for retrieving
programs and processing data. Storage consists of drives (hard, optical, USB, solid state).
Memory consists of RAM chips that lose their content when power is removed.”)
36
See, e.g., “THE DIFFERENCE BETWEEN MEMORY AND STORAGE” available at
http://www.technick.net/public/code/cp_dpage.php?aiocp_dp=guide_umg_01_003 (“People often
confuse the terms memory and storage, especially when describing the amount they have of each.
The term memory refers to the amount of RAM installed in the computer, whereas the term
storage refers to the capacity of the computer’s hard disk.”); “What's the Difference Between
Memory and Storage?,” available at http://www.tucows.com/article/593.
37
Vermaat, M., Sebok, S., Freund, S., Campbell, J. and Frydenberg, M., “Discovering
Computers” Cengage Learning (2016).
38
Opp. at 25:2-4.
34

- 15 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

1

45.

2

stated that

My testimony cited is taken grossly out of context in two ways. First, I never

3
It was never clarified.39 It is my opinion that

4
5

.

6

46.

Secondly, I opined that

7
8
9
10
42

11
12
13
14

Dated: February 19, 2016

15
16

__________________

17

Jennifer Golbeck

18
19
20
21
22
23
24
25

39
40

App. 1096–1101 (Golbeck Depo. Tr. at 103:13-108:16).
App. 1209-1210 (Golbeck Depo. Tr. at 216:7-217:5)

42

App. 1210-1216 (Golbeck Depo. Tr. at 217:6-223:6).
App. 1209-1212 (Golbeck Depo. Tr. at 216:7-219:3).

26
27
28

- 16 -

REBUTTAL REPORT OF DR. GOLBECK IN SUPPORT
OF PLAINTIFFS’ MOTION FOR CLASS
CERTIFICATION C 13-05996 PJH (MEJ)

Appendix A: List of Materials Relied On
I relied on the following documents and materials in forming my opinions:
Documents from Campbell et al. v. Facebook, Inc.:
Facebook’s Second Supplemental Responses and Objections to Plaintiffs’ Narrowed Second Set
of Interrogatories, and Exhibit A thereto
Expert Report of Dr. Benjamin Goldberg submitted with Defendant Facebook Inc.’s Opposition
to Plaintiff’s Motion for Class Certification
Declaration of Alex Himel in Support of Defendant Facebook, Inc.’s Opposition to Plaintiffs’
Motion for Class Certification
Declaration of Dan Fechete in Support of Defendant Facebook, Inc.’s Opposition to Plaintiffs’
Motion for Class Certification
Defendant Facebook, Inc.’s Opposition to Plaintiffs’ Motion for Class Certification
Deposition of Dr. Jennifer Golbeck (Dec. 16, 2015)
Deposition of Dr. Benjamin Goldberg (Feb. 2, 2016)
Report of Dr. Jennifer Golbeck in Support of Plaintiffs’ Motion for Class Certification
Plaintiffs’ Motion for Class Certification
FB000005575
FB000005577
FB000005800
FB000005882
FB000006007
FB000006088
FB000012006
FB000012557
FB000014213
FB000027011
FB000027015
FB000027018
FB000027190
FB000027191
Source Code Produced by Facebook

Other Materials:
https://www.facebook.com/business/help/133609753380850
Hasin Hayder, Object-Oriented Programming with PHP5 (Packt Publishing Ltd, 2007)
www.google.com (search for “Volkswagen ‘software device’”)
“EPA, California Notify Volkswagen of Clean Air Act Violations / Carmaker allegedly used
software that circumvents emissions testing for certain air pollutants,” available at
http://yosemite.epa.gov/opa/admpress.nsf/a883dc3da7094f97852572a00065d7d8/dfc8e33b5ab16
2b985257ec40057813b!OpenDocument
U.S Patent No. 6,032,223
V. Kreinovich, A. Bernat, E. Villa, Y. Mariscal, “Parallel computers estimate errors caused by
imprecise data,” Interval Computations, 1991, No. 2, pp. 31–46. (available at
http://ntrs.nasa.gov/search.jsp?R=19930068753)
http://windows.microsoft.com/en-us/windows7/memory-and-storage
http://www.pcmag.com/encyclopedia/term/63352/storage-vs-memory
http://www.technick.net/public/code/cp_dpage.php?aiocp_dp=guide_umg_01_003
http://www.tucows.com/article/593
Vermaat, M., Sebok, S., Freund, S., Campbell, J. and Frydenberg, M., “Discovering Computers”
Cengage Learning (2016).