White v. Crook Deposition of John Kraft

    APPENDIX B

Deposition of John F. Kraft



UNITED STATES DISTRICT COURT 
MIDDLE DISTRICT OF ALABAMA 

NORTHERN DIVISION

x

GARDENIA WHITE, et al.,

Plaintiffs,

UNITED STATES OF AMERICA,

Intervenor-Plaintiff 

-against -

BRUCE CROOK, et al. ,

Defendants

Civil Action 

No. 2263 -N

x

Deposition of JOHN F. KRAFT as a witness 

taken by the plaintiffs pursuant to stipulation, 

at 527 Madison Avenue (Room 601) , New York, New 

York, on Wednesday, November 24, 1965, at 

three-forty-five o'clock in the afternoon, before 

Marvin Slotoroff, a Notary Public within and for 

the State of New York.

BAR  A S S O C IA T IO N  R E P O R T IN G  S E R V IC E  
W ILLIA M  N E W R O C K . C S R

3 6  W E S T  4 4 tm  S T R E E T  

N E W  Y O R K  SS. N . Y .

M URRAY Mi l l  7 -0 9 8 5



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

u

P R E S E N T :

MELVIN L. WULF, ESQ.
Attorneys for Plaintiffs 
American Civil Liberties Union 
156 Fifth Avenue 
New York, New York

MR. WULF: Let the record show that this

is the deposition of Mr. John F. Kraft taken on 

behalf of the plaintiffs Gardenia White, et al., 

in the suit entitled Gardenia White, et al., 

Plaintiffs, United States of America, Intervenor- 

Plaintiff, against Bruce Crook, et al.,

Defendants, Civil Action No. 2263-N in the United 

States District Court, Middle District of Alabama, 

Northern Division.

Let the record also show that the attorneys 

for each of the defendants have waived notice 

and all rights to appear and be present at this 

deposition, and have waived all objections to 

the form of the questions, and have waived the 

witness's signature.

B A R  A S S O C IA T I O N  R E P O R T IN O  A N D  S T E N O G R A P H IC  S E R V IC E
9S  W E S T  44 TH  S T R E E T .  N E W  Y O R K  3S. N . Y ., M URRAY H lU .  7 -0 3 8 5



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

3

J O H N  F. K R A F T ,  having

been first duly sworn by the Notary Public, was 

examined and testified as follows:

EXAMINATION BY MR. WULF:

Q Would you state your name, please?

A John F. Kraft.

Q And your business address?

A 527 Madison Avenue, New York City.

Q Would you state your occupation, please?

A I am the president of an opinion and market research

firm.

Q Would you state your educational background,

please?

A I have a BA from Yale University and I did graduate

work at Columbia University for one year.

Q When did you get your undergratuate degree?

A In 1949.

Q IIow long have you been in the polling pro­

fession?

A Fifteen and one-half years.

Q Would you relate your professional history?

A My first six years in this business were spent

with Elmo Roper 5 Associates.

The next three years were as a partner in Louis

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
9 0  W E S T  A4TM S T R E E T .  N E W  Y O R K  3 « .  N . Y .. M URRAY H IL L  7 -O SSS



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

Kraft 4

Harris § Associates.

And for the last six and one-half years, I have 

headed up my own firm.

Q Would you describe the nature of the work in

which you have engaged during your being employed in this 

profession?

A It falls into two broad categories, one cf them 

being political research, which is that which is heard 

about most often.

The other and larger area is confidential consumer 

research for clients in the business area.

In the six and one-half years we have been in 

business, we have, for example, conducted a number of 

research projects for the Union Carbide Chemicals 

Company, a number of studies for the Union Carbide Con­

sumer Products Company, several projects for the Corning 

Glass Works, a number of studies for the Ford Motor 

Company, and through N. W. Ayer, studies for such 

clients of theirs as Armour, Johnson § Johnson, Illinois 

Bell; for BBD § 0, a study for Air France; for Geyer, 

Morey 5 Ballard, some studies for Rambler, and also for 

many others.

Q Would you say for whom you have worked in

your political polling work?

B A R  A S S O C IA T I O N  R E P O R T IN O  A N D  S T E N O G R A P H IC  S E R V IC E
a e  W E S T  4 4 t h  S T R E E T ,  n e w  Y O R K  a s . N . Y .. M URRAY h i l l  7 -o a s s



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 5

A The political research we are doing now is, of

course, confidential.

But I can say that in the past, we have conducted 

political research for such senators as Carl Hayden,

George Symington, George Smathers, William Fulbright,

Ross Bass, Lee Metcalf, Quentin Burdick, and others.

Q Could you describe generally the purpose of

one or two of the polls or sample projects that you have 

undertaken?

A Commercial or political?

Q One or two of each.

A All of our work has required our getting representa­

tive cross sections of the people with whom a client 

might be concerned.

For example, for Ford, we have sought representative 

cross sections of purchasers of new Ford automobiles.

For Corning Glass, we have had to select 

representative cross sections of housewives.

And in many media studies we have had to select 

representative cross sections of subscribers to particular 

magazines.

In the political area, any study we undertake must 

be based on the proper selection of a cross section of 

the voting population, whether it be a study in Wyoming

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
S «  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  3 « .  N . Y .. M URRAY H i l l  7 -0 3 8 8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 6

or Georgia or New York City.

Q Can you briefly describe the method by which

you undertake a poll, whether it's for a political or 

commercial client? I am interested in the techniques 

involved in obtaining the information necessary for 

the particular poll involved.

Q If we were doing a study among subscribers

to a magazine, we would request listings of subscribers 

on a county by county basis throughout the United 

States or on a state by state basis, depending on the 

need, and select every one-hundredth, every tenth, or 

even every one - thousandth subscriber for interviewing 

purposes.

There are other ways to approach a random selection 

of subscribers, but this should suggest a very basic 

approach.

In a political survey, we would allot clusters of 

interviews within a state on a random basis proportionate 

to the description of the population within that state.

To clarify, we could decide that 100 interviewing 

points at which ten interviews were to be done -- at 

each of which ten interviews were to be done, that is -- 

would produce an accurate cross section of the entire 

voting population.
B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  36 . N . Y .. M URRAY H i l l  7 -0 3 6 8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 7

We would then allocate those interviewing points 

of ten interviews each to counties within a state at 

random and porportionate to the size of those counties.

Now, if more than one point were to fall within 

a larger county, we would then allocate these points at 

random, proportionate to the size of population within 

the county.

At a given interviewing point, we would specify 

to an interviewer that he or she must begin at a given 

point -- that is, a given point which had previously 

been randomly selected -- and he would be directed 

to begin at the third household at the northeast corner, 

for instance, interviewing at every fifth household 

around the block or down the street or all along the 

rural route -- whichever was relevant to the local 

situation.

Q What is the degree of accuracy with respect

to the entire population with which you are concerned 

of the information which you obtain through investigation 

of randomly selected cross sections of the population?

A The accuracy can vary, depending upon how intensively

you sample a particular population or universe. That 

is to say, a national cross section of 2000 interviews 

on such a question as, "Would you vote for Johnson
B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
9 6  W E S T  44TH  S T R E E T .  H E W  Y O R K  36 . N . Y .. M URRAY H lU .  7 -0 3 8 8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 8

or Goldwater?" could be expected to produce an accurate 

prediction, so to speak, within 2 or 3 per cent.

This has been the kind of experience that Gallup 

has demonstrated during the last several presidential 

elections.

A more intensive sample, that is to say, sampling 

of one out of every ten people rather than one out of 

every ten million people, will, of course, be much 

more accurate.

There is, however, another factor to be considered 

and that is that if a preference, as in the Goldwater- 

Johnson illustration, is close to fifty-fifty, then 

the accuracy would be less great than it would be if a 

preference were ninety to ten, say.

However, there is no magic in doubling, tripling or 

quintupling the sample size in the sense that accuracy 

does not increase in direct proportion to the increase 

in sample size.

I should add that to talk about percentage accuracies 

I am doing this only very generally because accuracy 

is based on the size of the universe or population being 

considered and the intensity with which it is sampled, 

and also the complexity of that population or universe.

Q In your professional opinion and experience,
B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
SB W E S T  4ATM S T R E E T .  N E W  Y O R K  SB. N . Y .. M URRAY H IL L  7 -0 3 0 3



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft y

Mr. Kraft, have you been satisfied that the results of 

random sampling techniques are, in fact, very accurately 

reflective of the particular kind of information which 

you may be seeking in a particular poll?

A Not only are we satisfied but we have found, in hun­

dreds of studies, that the sampling techniques I have 

been describing here are really the only way to do a 

proper job of producing the kind of cross section that 

is needed in one situation or another.

Q I take it that, given these scientific random

sampling techniques, it is possible to investigate any 

particular cross section of the population which would be 

relevant to the study you are conducting?

A Yes.

Q What is the adaptability of the random sampling

technique to a different variety of problems presented 

to you by anyone who might employ you?

To be more specific, if anybody wants to know 

anything about a particular product or a particular 

opinion in which they are interested, are you able to 

provide this information to them accurately through the 

random sampling technique?

A Yes. All that is required is knowledge of what

the population requirements are for that particular
B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3S  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  3S . N . Y ., M URRAY H IL L  7 -O SSB



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 10

product.

As I noted earlier, it may be housewives, it may 

be people who smoke Philip Morris, it may be eligible 

voters in the State of Washington, it may be beer drinkers 

in New York City.

But once you understand the population you are 

dealing with, then you can, and we have -- our colleagues 

and competitors in this business have -- demonstrated 

time and again the reliability of proper random sampling 

of cross sections of the population on a scientific 

basis.

Q Then you can provide any information of any

sort, just so long as it has to do with human behavior?

A That is right. Some are more difficult than

others.

Q To take a bizarre example, if you, sir,

wanted to know, for instance, how many people who owned 

green cats would be interested in buying a narticular 

brand of cat food, would you be able to establish 

something of that sort?

A Assuming that there were green cats to be found,

proper random sampling methods could determine not only 

how many green cats there were but also the preferences 

for particular cat food as this affected the green cats

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  36 . N . Y .. M URRAY H i l l  7 -0 3 8 5



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 11

0 £0 you would describe, then, the random sampling

technique as a precise and scientific method for obtain­

ing information of a great variety of sorts?

A It's indeed precise, within some of the limits I

have suggested -- the one or 2 or 3 per cent error 

and it'sscientific because it removes the human element 

from deciding whom to interview or what the universe may 

be. It is impersonal.

Q If I put to you an example of this kind: Let

us assume there is a county in the State of Alabama 

which has a population of approximately 15,000, which 

consists of a Negro population of approximately 12,500 

and a white population of 3000, and assume further that, 

as in any representative community, there is not only 

the description of population by race, but also by age, 

education and economic status. And if I were to ask you 

whether it would be possible to apply random sampling 

techniques in order to obtain a very accurate cross 

section of this particular community, in order to assure 

that an accurate cross section of the community were 

available for service on juries, would the techniques you 

have described be applicable to this problem?

A There is no reason why they could not or would not

be .
B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  36 . N . Y .. M URRAY H IL L  7 -0 3 6 8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 12

0 Do you mean that there isn't any question

that it could be?

A I mean there is no question that it could be done.

Q In other words, the technique is trans­

ferable to this particular situation?

A Not only that, but it would be easier than some 

of the problems I have described to you earlier'.

Q Why do you say that?

A The information on the population of a county

in Alabama is much more complete in terms of sex, age, 

economic level, race, average rental and other kinds 

of information than it is for any other universe or kind 

of population I have discussed before.

There is really no census material on who used 

Pyrex. There is inadequate information on who owns new 

Ramblers, and there is not much information on who switched 

from Marlboros to L § Ms last week.

Q But the kind of characteristics which I have

described are -- I assume what you are saying is that 

they are readily identifiable?

A That is right.

(continued on next page)

B A R  A S S O C IA T I O N  R E P O R T IN O  A N D  S T E N O G R A P H IC  S E R V IC Eas W E S T  44 TH  S T R E E T .  N E W  Y O R K  3 6 . N . Y .. M URRAY H i l l  7 -O SSS



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 13

0 Could you describe how the random sampling 

technique could be adapted to the selection within 

Lowndes County, for example, of a cross-sectional list 

of prospective jurors?

A I would assume that the kind of county you are

talking about, as would be true of any other county, 

would have the same Bureau of the Census information 

on enumeration districts, census tract or block statistics 

as any census would.

0 You are referring, I assume, to the Federal 

decennial census.

A That's right. And in using this information, you 

should select at random and‘impartially a relevant 

number of enumeration districts which would be designed 

to provide the number of jurors which might be needed.

For example, in that county you might pick forty 

blocks, or forty enumeration districts, depending upon 

the characteristics of the county, and within each of 

those randomly selected blocks do what in effect would 

be a census of every fifth household in those blocks, 

and within those households make a random selection of 

citizens eligible to serve on juries -- that is to say, 

men and women over 21, upper income and lower income 

people, whites and Negroes, and so on.

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
9 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  96 . N . Y ., M URRAY H IL L  7 -0 9 8 9



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 14

Q And it wouldn’t affect your judgment, would it, 

if, as in Alabama, women are restricted from jury service? 

Your technique would be adapted to that fact, would it 

not?

A Well, there would be two approaches we would 

consider. We could automatically eliminate, again, in 

scientific fashion, the women from such a sampling or, 

on the other.hand, they could be included in a sampling, 

which would necessarily be a larger sampling, in 

anticipation of most of these women expressing a disinclinp 

tion to serve on a jury.

0 Would it be very difficult to set up the 

mechanics of operating a jury selection system on a random 

sampling basis?

A I assume you mean from the point of view of a 

local --

0 Could a local resident of Lowndes County, of 

common intelligence and average education, do whatever 

would be necessary in order to set up and, administer 

such a jury selection system?

A Yes, but with this qualification:

The sampling approach T have been talking about 

would have to be summarized for that local person 

to put into action.
B A N  A S S O C IA T I O N  N E P O R T IN O  A N D  S T E N O G R A P H IC  S E R V IC E
9 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  96 . N . Y .. M URRAY H IL L  7 -0 9 6 9



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 15

The summarization could be simple and quite 

straightforward, perhaps adding up to a page or two or 

three of directions specifying such things as, "Go to 

the corner of Maple and Ash Streets and get the names of 

the people in every third household", as I have described 

the procedure earlier.

0 So this initial preparation would have to be

done by experts, but it could be implemented with 

easily-understood instructions?

A Yes, but the experts would not have to take a

great deal of time to set up such a system.

For someone trained in the business of sample 

selection, it would be a relatively straightforward and 

simple operation provided the kind of census material 

I have talked about were made available.

It could be done, for example, I am sure, by people 

in the Bureau of Labor Statistics within a state, or 

a Chamber of Commerce, or comparable organizations with 

census material available.

0 Do you know of any other system of selection 

which might achieve the same accurate cross-sectional 

results in the jury selection process as the random 

sampling technique you have described?

A No. There is none that we have heard of and none

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 0  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  30 . N . Y .. M UR RAY HILL. 7 -0 3 0 9



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 16

that we have been able to discover.

Q As you no doubt know, there are a variety 

of selection techniques used in drawing jurors around 

the country, not including the random sampling techniques 

but rather using such devices as telephone directories 

or property tax rolls or gas and electric company 

records or voting records.

Do you have an opinion about the cross- 

sectional accuracies of those kinds of lists?

A Yes. And let me take up a couple of them.

You mentioned telephone listings. In some sections 

of the country, for example in New York City, there are 

as many as 400,000 unlisted telephones, so that any 

survey sample is going to omit representation of the 

kinds of people who do not want their telephones listed 

in New York City.

For quite different reasons, there are many 

unlisted telephone numbers in a city like Houston.

Secondly, where telephones are concerned, this is 

an expense,so that resorting to telephone listings would 

automatically exclude from any sound sampling procedure 

those people who are too poor, or consider themselves to 

be too poor, to have a telephone in the home.

In other words, you would have a bias in the

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
S S  W E S T  4 4 TH  S T R E E T .  N E W  Y O R K  3 8 . N . Y .. M URRAY H lU .  7 -0 3 8 8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 17

direction of upper income people by going to the 

telephone listings in this situation.

Taking the example of the utilities, in many rural 

areas there are a great many people who still rely on 

coal, kerosense, candlelight and just plain daylight, 

and resorting to such lists would, again, automatically 

exclude people who were too poor, or too far down the 

economic ladder, to afford utility service.

You mentioned voter lists. There are wild 

variations in this area because a great many people in 

a great many counties are unregistered, or just do not 

vote or, in fact, in counties or states where there is 

a large transient population, inevitably an effort to use 

voting lists may omit as many as 15 to 25 per cent of 

the resident population.

You also mentioned using property tax rolls. 

Again, this strikes against lower income people, whether 

they be people with large families, people who are 

sporadically employed, or what have you.

0 Is there any kind of list, except census 

material, which can be used to get a representative 

c t o s s -section of a given population?

A Not if you are undertaking to do a survey, for 

example, for a political client, or for a client who has

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
a s  W E S T  4 4 t h  S T R E E T . N E W  Y O R K  S «. N . Y .. M UR RAY H il l . 7-03S8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 18

a product which is in general use throughout a county, 

a state or a region.

You can, of course, go to specific lists where you 

have a very narrowly defined population, such as 

subscribers to a magazine, or new car owners, as I 

suggested earlier, but that would be doing a disservice 

to the client and guaranteeing, in fact, inaccuracies 

to do a survey, for example, for a political client from 

some of the lists we have been discussing here, because 

we would be guaranteeing the omission of key groups of 

people, such as women, people who didn't own property, 

people who didn't own telephones, people who preferred 

kerosene to electric power, and so forth.

0 In your opinion, Mr. Kraft, what would be 

the cross-sectional accuracy of a list of jurors 

selected by jury commissioners, for example, on the 

basis of their personal acquaintances or by their 

writing large numbers of letters to different groups of 

people, or even to individuals within a community?

A Well, if they were to communicate with their 

friends and acquaintances, they just conceivably might be 

reproducing a cross-section of their friends and 

acquaintances, but they certainly wouldn't be reproducing 

a cross-section of a total population in an area because

E A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
M  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  3 6 . N . Y ., M URRAY H IL L  7 -0 3 8 0



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft

it is an incredibly rare person who has an acquaintance­

ship with a representative cross-section of an entire 

population.

With regard to writing letters to large numbers of 

people, first we have to face the problem of whether or 

not mail delivery is available to all of the people in 

the population, whether addresses are correct and 

whether in fact his signature will spur or discourage 

response to his letter.

So the two examples you ask about seem to me to 

insure unrepresentativeness, quite the opposite from 

a representative cross-section.

0 In the random sampling selection technique 

of jurors which you have described,is there any room 

whatsoever for the exercise of subjective discretion in 

the selection of jurors by the persons administering the 

selecton process?

That is to say, do the personal preferences of 

the person administering this random sample jury selection 

technique enter into the selection process at all?

A Not the process we have been describing, not in any 

way.

Let me explain it this way: The whole focus of

our approach is to remove human discretion from

19

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
36 W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  36. N . Y .. M URRAY H i l l  7 -0 3 6 5



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 20
scientific sampling, to remove the possibility of 

bias in selecting the sample, and the kind of sample 

selection I have been describing here is one which will 

effectively do that, and I believe it is really the only 

way to insure selecting a random representative cross- 

section of a given population and be totally unbiased.

0 In conclusion, do you know of any system

other than the random sampling system based upon 

Federal census material that can produce the same highly 

accurate cross-section of a population?

A There would be no other material available as of 

right now to produce a better approach.

There is another system, though, which might be 

employed, which would in effect add up to a re-census.

I will explain that last by saying that you could impose 

a geographic grid system on a state or country or county 

or town you are concerned about, and then select squares 

from that grid, and if you will visualize a grid system 

of let's say 100 by 100 squares, for a total of 10,000 

squares, in other words, you could select at random 

some 100 squares and do a census within each one of 

those squares of the adult population,and then apply 

similar methods to those that I have described earlier, 

which is to say, a random sampling of people within each

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
36 W E S T  44 TH  S TR E E T . N E W  Y O R K  36. N . Y .. M URRAY H IL L  7 -0 3 8 5



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft 21

of those squares proportionate to the distribution of 

the population.

0 Doesn't that in effect come to the same

thing as the random sampling technique?

A In effect, yes, it does.

It is more complicated and more complex, and it is 

a harder way to do it, and you would only employ it in 

the absence of adequate census material.

0 Do you have any douht about the feasibility

and the practicality of applying random sampling 

techniques to the jury selection process?

A No, I don't, if it would be understood that a

random sampling technique in the county you mentioned 

earlier with the population of 15,000 were to be based 

on, for example, a cross-section of 1,000 potential 

members of a panel -- well, it is possible that we might 

under-represent or over-represent upper income people 

to the tune of one per cent or two per cent, and 

similarly, we might over or under-represent people with 

large families by one or two per cent -- but if this 

margin of error on that kind of a sample is acceptable, 

which it is in my opinion, then there is absolutely 

no question that this is the sound and proper approach.

0 You wouldn't think that would be a critical

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  36 . N . Y .. M URRAY H IL L  7 -0 3 8 8



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Kraft
22

variation in this kind of problem, in other words? 

A By no means.

MR. WULF: I believe those are all the

questions I have.

(Whereupon at 4:30 P.M. the deposition 

concluded.)
* * * *

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 6  W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  36. N . Y .. M URRAY H IL L  7 -0 3 S S

w n c



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

2 3

NOTARIAL CERTIFICATE

STATE OF NEW YORK )
) SS:

COUNTY OF NEW YORK)

I, Marvin Slotoroff, a Notary Public 

of the State of New York, do hereby certify 

that the foregoing deposition of JOHN F.

KRAFT was taken on behalf of the plaintiffs 

in the above entitled action, before me at 

Room 601, 527 Madison Avenue, New York, New 

York, on Wednesday, November 24, 1965, 

commencing at three - forty-five o'clock in the 

afternoon, and concluding on the same day.

I do further certify that the said 

witness was by me duly sworn before the 

commencement of his testimony; that the 

testimony of the said witness was taken 

stenographically by myself and then transcribed: 

that the within deposition is a true and 

accurate transcript of my stenographic notes; 

and that the parties were represented by counsel 

as indicated on Page 2 of the within deposition.

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
3 S  W E S T  44TH S T R E E T .  N E W  Y O R K  3 « .  N . Y .. M U R R A Y  H i l l  7 -O S8S



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

I do further certify that I am not 

an employee of, nor connected by blood or 

marriage with, any of the parties hereto or 

their counsel, nor interested directly or 

indirectly in the matter in controversy.

IN WITNESS WHEREOF I have hereunto 

set my hand and seal this 24th day of 

November 1965.

Notary Public of the State of New 
York, residing in New York Countv 
My commission expires March 30,13

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
30 W E S T  44TH S T R E E T .  N E W  Y O R K  36. N. Y.. M U R R A Y  Hlt-L 7-03B«



L A W Y E R ’S N O T E  S
P a g e L i n t *

.

B A R  A S S O C IA T I O N  R E P O R T IN G  A N D  S T E N O G R A P H IC  S E R V IC E
BS W E S T  4 4 t h  S T R E E T .  N E W  Y O R K  3S. N . Y ., M UR RAY H i l l  7 -0 3 3 8