characteristics of response tendency in mail surveys · gave the following summary regarding middle...

1

Characteristics of Response Tendency in Mail Surveys

Comparing Mail and Face-to-Face Surveys

Public Opinion Research Division Hiroko Murata Noriko Onodera

1. Introduction

This paper focuses on a comparative review regarding the distribution of responses of a mail

survey and a face-to-face interview1), both conducted in 2010.

In 2008, the NHK Broadcasting Culture Research Institute started the Survey Methodology

Project and performed a comparative study on survey methodology. In June 2008, an experimental

survey was conducted in accordance with the 8th Survey on Japanese Value Orientations 2008. The

experimental survey reviewed different data collection methods (face-to-face interview,

drop-off/pick-up survey and mail survey) and compared the distribution of responses between the

survey modes. The comparative study provided beneficial information regarding face-to-face

interviews, as well as clues for standardizing methods for mail surveys. There was a difference of

responses between the survey modes, with a particularly large difference between self-administered

modes (drop-off/pick-up survey and mail survey) and interviewer-administered modes (face-to-face

interviews). However, the difference of responses for each mode was thought to be caused by

multiple factors such as social desirability, resistance to response, and the contents of the

questions2).

In this way, it can be inferred that the difference of responses between survey modes varies

according to factors such as survey topic, content and wording of questions, as well as arrangement

of response options. It is necessary to continue to accumulate various data in order to examine mode

effects.

In addition to research like the Survey on Japanese Value Orientations which probes social

attitudes, NHK also conducts numerous attitude surveys regarding broadcasting. In March 2010,

NHK conducted an experimental mail survey which contained the same questions as The Japanese

and Television 2010 Survey (face-to-face interview)3), a survey which tracks popular attitudes and

behavior toward television. We then compared the distribution of responses.

Furthermore, we decided to review the effects of middle options for the self-administered

questionnaire. In the recent The Japanese and Television 2010 Survey, a face-to-face interview

which was the subject of the comparison, there were many questions containing responses which

could be interpreted as middle options, such as “Neither yes nor no.”

According to Yamada (2010), middle options encompass a wide variety of responses including

December, 2011

2

1) situations which can neither be affirmed nor denied, 2) the expression of ambiguous attitudes in

which any response is applicable, and 3) situations in which the respondent has never thought of the

question or cannot find an answer even after consideration.

Moreover, from a comparison of face-to-face interviews and the mail survey, Maeda (2005)

gave the following summary regarding middle options: “Since mail surveys are self-administered

surveys, middle options attract a much larger portion of respondents. Thus, merely not providing

response options is not enough to conduct a comparative review between mail surveys and

face-to-face interviews.

Therefore, in order to conduct a comparative review for cases in which middle options are

either provided or not provided, the experimental survey for the mail mode was conducted using a

split ballot method (a method in which respondents are randomly divided and distribution of

responses are compared).

A summary of the surveys is shown in Table 1.

Table 1: Overview of Experimental Survey

Experimental Survey (Mail) Face-to-Face Interview

Survey title The Japanese and

Television 2010 Survey The Japanese and

Television 2010 Survey Survey method Mail survey Face-to-face interview

Survey period March 3rd (Wed.)

to April 16th (Fri.), 2010 March 6th (Sat.)

& 7th (Sun.), 2010 Sampling frame Basic Resident Register

Sampling method Stratified two-stage random sampling Population Japanese aged 16 and over Sample size 1,800(12 people×150 units) 1,800(12 people×150 units)

The number of respondents

(response rate)

Mail All

1,046(58.1%) Mail A Mail B

1,215 (67.5%)

609 (67.7%)

606 (67.3%)

In the remainder of this report, The Japanese and Television 2010 Survey conducted using

face-to-face interviews shall be abbreviated as the “Face-to-Face”, and the experimental survey

using mail shall be abbreviated as “Mail All.” Furthermore, for questionnaires divided using the

split ballot method, results for questionnaires providing middle options shall be referred to as “Mail

A,” while results for questionnaires not providing such options shall be referred to as “Mail B.”

For the Mail All, systematic sampling was performed for 12 respondents from each of 150

units throughout Japan. Additionally, in order to compare the provision/non-provision of middle

options, allocation was performed to achieve homogeneity among respondents. Allocation was

performed for 900 respondents in Mail A, with odd numbers assigned to respondents in

odd-numbered units and even numbers to respondents in even-numbered units. Similarly, allocation

was performed for 900 respondents in Mail B, with even numbers assigned to respondents in

odd-numbered units and odd numbers to respondents in even-numbered units.

3

Based on survey design for experimental surveys as previously mentioned, this report reviews

overall differences between survey modes, the difference of responses for mail surveys and

face-to-face interviews, and the difference of responses due to the provision/non-provision of

middle options.

The marginal count and sample distribution for Mail All are listed on pages 23 to 41 at the end

of this report. Marginal count for Face-to-Face are listed together with past results in the August

issue of The NHK Monthly Report on Broadcast Research.

2. Response Rate and Sample Distribution

（1）Response Rate

Figure 1 shows the response rate for each survey mode by gender and age groups. Overall, the

response rate is higher for Mail All than for Face-to-Face. Furthermore, when examining gender

and age groups, the response rate is higher for Mail All than for Face-to-Face in the case of young

age groups of both men and women. It can be inferred that the mail survey is an effective method

for young age groups in which the face-to-face interview survey does not often lead to valid results.

For older age groups, there was no difference in the response rate between different modes for both

men and women.

Figure 1: Response Rate by Survey Mode (by male/female age group)

（2）Sample Distribution

Table 2 shows the distribution of issued samples and valid samples for Mail All and

Face-to-Face. In the case of issued samples, both Face-to-Face and Mail All are in close agreement

with the population distribution of the Basic Resident Register in terms of gender, age groups, size

of city, and region as extracted from the Population Summary of Basic Resident Register (2009

version).

55 61

44

61

71

42

58

70

47

63

71 61

73

60

70 74

54 60

71 65

79 77

0

20

40

60

80

100

Men Women Young(16-39)

Middle (40-59)

Upper(60 andolder)

Young Middle(Men)

Upper Young Middle(Women)

Upper

Face-to-Face Mail All

（%）

4

Table 2: A Comparison of the Distribution of the Basic Resident Register and Each Sample

Total No. (people)

Gender and age group (%) Men Women

Young (16-39)

Middle (40-59)

Upper(60 and older)

Young (16-39)

Middle (40-59)

Upper(60 and older)

Basic Resident Register

108,661,394 17.8 15.5 15.2 17.0 15.3 19.2

Issued samples

Mail All 1,800 17.4 15.3 15.5 18.8 15.3 17.8 Mail A 900 18.1 14.4 14.9 18.6 16.3 17.7 Mail B 900 16.7 16.1 16.1 19.0 14.2 17.9 Face-to-Face 1,800 18.7 14.7 15.1 17.3 15.9 18.2

Valid samples

Mail All 1,215 13.9 13.5 16.3 18.0 17.9 20.3 Mail A 609 14.0 14.1 15.4 17.9 18.4 20.2 Mail B 606 13.9 12.9 17.2 18.2 17.5 20.5 Face-to-Face 1,046 13.4 14.7 18.2 14.1 17.3 22.3

Total No. (people)

Size of city (%)Special wards

& cities of 1 million or

more

Cities of 300 thousand or

more

Cities of 100 thousand or

more

Municipalities of

50 thousand or more

Municipalities of less than 50 thousand


127,076,183 21.8 22.2 24.0 14.9 17.1

Issued samples

Mail All 1,800 20.7 22.7 24.0 15.3 17.3 Mail A 900 20.7 22.7 24.0 15.3 17.3 Mail B 900 20.7 22.7 24.0 15.3 17.3 Face-to-Face 1,800 21.3 22.0 23.3 17.3 16.0

Valid samples


Total No. (people)

Region (%)

Hokkaido/Tohoku

Kanto Chubu/

KoshinetsuKinki

Chugoku/ Shikoku/ Kyushu


127,076,183 11.8 32.7 18.5 16.3 20.7

Issued samples


Valid samples


* The Population Summary of Basic Resident Register is calculated for age groups in 5-year intervals. Therefore, the value used for the teen age group (16-19 years) was calculated by multiplying the total for 15-19 years by 4/5.

* At the time of sampling, the total population number was used to assume the size of cities and regions. Therefore, the composition for the Basic Resident Register was also calculated using total population.

5

However, valid samples in Mail All, Mail B, and Face-to-Face show a distribution different

from that of the Basic Resident Register for gender and age groups. As a result, distortion can be

observed in the sample. Mail A was in close agreement with the distribution of the Basic Resident

Register4).

Next, valid samples of the experimental survey and Face-to-Face were examined to verify

differences in the population distribution of gender and age groups, size of city and region. Results

for χ2 test (test of independence) showed no statistical significance for either survey, and no

differences in distribution were observed.

3. Overview of Differences in Survey Results

（1）Overview of Differences in Responses between Survey Modes

First, we will examine overall trends of differences in results between survey modes. Response

trends differ for Mail A, which provided middle options, and Mail B, which did not provide such

responses. Therefore, each experimental survey was compared separately with Face-to-Face.

Analysis was performed for 321 response options of 74 questions, excluding nonresponses and

middle options. Specifically, the following types of response options were excluded from the

analysis.

・Nonresponse items which were not listed in the questionnaire for Mail All or in the showcard for

Face-to-Face.

・Response options “Neither yes nor no,” (“Neither of the above,”) “None of the responses apply,”

“Not listed here,” and “Other” which were provided only for Mail A. (Responses which were also

provided for Mail B were not excluded.)

Additionally, questions 19A, B, C, D, E, F and G were excluded because many respondents

selected multiple answers despite the fact that these were single answer questions for Mail All5).

For applicable response options, a correlation coefficient and scatter diagram (Figure 2) were

used to confirm the degree of agreement between responses (percentage) for Mail A, Mail B and

Face-to-Face. A correlation coefficient of 0.976 was confirmed when comparing Face-to-Face and

Mail A, which provided middle options. A correlation coefficient of 0.975 was confirmed when

comparing Face-to-Face and Mail B, which did not provide such response options. Correlation is

high because each survey mode used the same questions.

Additionally, when including the middle option “Neither yes nor no” in the analysis, a

correlation coefficient of 0.965 is confirmed when comparing Mail A and Face-to-Face.

6

Figure 2: Comparison of Responses by Survey Mode

To continue, when examining the distribution of response differences (absolute value) between

survey modes (Table 3), 42 percent of the responses have a response difference of less than 2

percent when comparing Mail A and Face-to-Face. Similarly, when comparing Mail B and

Face-to-Face, 45 percent of the responses have a response difference of less than 2 percent. As

stated previously, although there is high correlation between each survey mode, the response

difference of less than 2 percent is limited to approximately 42-45 percent for both experimental

surveys.

Table 3: Distribution of Absolute Values for Response Differences (all questions)

Absolute value for response (%)

difference

Face-to-Face – Mail A Face-to-Face – Mail B No. of

response options

% Cumulative

%

No. of response options

% Cumulative

%

Less than 2.0 135 42.1 42.1 144 44.9 44.9 2.0-3.9 70 21.8 63.9 68 21.2 66.1 4.0-5.9 53 16.5 80.4 37 11.5 77.6 6.0-7.9 28 8.7 89.1 29 9.0 86.6 8.0-9.9 20 6.2 95.3 22 6.9 93.5

10.0 or greater 15 4.7 100.0 21 6.5 100.0 Total 321 100.0 321 100.0

(%)

(%)

Mail A

Mail B

Face-to-Face

Face-to-Face

7

As shown in the scatter diagram, the magnitude of correlation for response differences

excluding middle options is as follows: Mail A<Face-to-Face<Mail B. This correlation exists

because Mail A provides middle options while Mail B does not provide such options (refer to p.13

regarding middle options).

Furthermore, we extracted only questions which are not affected by the existence/non-existence

of middle options and summarized such questions in Table 4. The subject for this table is 164

response options of 31 questions for which there was no difference in the response options provided

in Mail A and Mail B. When compared with the distribution for response difference among all

questions (Table 3), the differences in response for the extracted questions is relatively small. 52

percent of the responses have a response difference of less than 2 percent when comparing Mail All

and Face-to-Face. The difference is smaller in the case of questions for which response options are

provided under the same conditions. When comparing Mail A and Mail B, 71 percent of the

responses have a response difference of less than 2 percent. There was a small difference between

surveys using the same survey method.

Table 4: Distribution of Absolute Values for Response Differences (only for questions with the same

response options in Mail A and Mail B.


difference

Face-to-Face – Mail All Mail A – Mail B No. of

response options

% Cumulative

%


% Cumulative

%

Less than 2.0 86 52.4 52.4 116 70.7 70.7 2.0-3.9 29 17.7 70.1 30 18.3 89.0 4.0-5.9 29 17.7 87.8 14 8.5 97.6 6.0-7.9 14 8.5 96.3 4 2.4 100.0 8.0-9.9 5 3.0 99.4 0 0.0 100.0

10.0 or greater 1 0.6 100.0 0 0.0 100.0 Total 164 100.0 164 100.0

（2）Comparison of Nonresponses between Survey Modes

Regarding the same 74 questions which were first analyzed on the previous page, Table 5 is a

summary of the distribution for “don’t know, nonresponse” (percent) items which were not

provided on the questionnaire for the mail survey or the showcard of the face-to-face interview.

Table 5: Distribution of Nonresponses

Non- response rate (%)

Mail A Mail B Face-to-Face No. of

response options

% Cumulative

%


% Cumulative

%


% Cumulative

%

Less than 2.0

60 81.1 81.1 68 91.9 91.9 19 25.7 25.7

2.0－3.9 13 17.6 98.6 4 5.4 97.3 37 50.0 75.7 4.0－5.9 1 1.4 100.0 0 0.0 97.3 15 20.3 95.9 6.0－7.9 0 0.0 100.0 2 2.7 100.0 3 4.1 100.0

Total 74 100.0 74 100.0 74 100.0

8

Overall, when comparing Mail A and Mail B with Face-to-Face, there was a large number of

questions with a low nonresponse rate. The 2008 NHK Experimental Survey reported a low

nonresponse rate for self-administered surveys (drop-off/pick-up and mail) when compared to

face-to-face interviews. Similarly, our comparison confirmed a low nonresponse rate for the mail

survey.

Furthermore, when comparing Mail A and Mail B, items with a nonresponse rate of less than 2

percent were a particularly high 92 percent for Mail B, which did not provide middle options. Since

almost no questions provided middle options, respondents to Mail B may have felt strongly

compelled to select a response for each question. However, for the multiple-answer format Question

12, there were many nonresponses even for Mail B (Table 6).

Table 6: Questions with Large Number of Nonresponses Question

no. Question Title

Mail A %

Mail B %

Face-to- Face %

12B Evaluation of Programming (Commercial broadcasting) 3.1 6.4 7.9 30 No. of Commercial Broadcasting Channels 2.5 2.5 7.4 12A Evaluation of Programming (ＮＨＫ) 2.0 7.1 6.3 11B Image of Television Influence B: Standardization 2.5 0.5 5.9 18C Frequency of Access to Media C: HDD 4.1 2.5 5.9 21D Comparison of Useful Media D: Selection 1.3 1.2 5.6 11C Image of Television Influence C: Violence 2.0 0.3 5.0 11D Image of Television Influence D: Material or spiritual 1.8 0.7 5.0

* Shaded areas are questions with an item nonresponse rate of 5% or higher.

Table 6 shows questions with a high number of nonresponses for each survey mode. In the

case of Face-to-Face, questions for which it was difficult to answer the interviewer immediately

resulted in a high number of nonresponses. Such questions included the “Evaluation of

Programming,” in which the respondent is asked his/her impression of NHK and commercial

broadcasting, and the “Image of Television Influence,” which concerned the influence of television

on the daily lives and thoughts of people. There were also a high number of nonresponses for

questions which required a certain extent of prior knowledge. Such questions included the “No. of

Commercial Broadcasting Channels,” in which respondents were asked the number of commercial

broadcasting channels which they can watch at home, and the “Frequency of Access to Media C:

HDD,” which asked respondents how much they use hard disc recorders.

4. Difference in Results between Mail Survey and Face-to-face Interviews

Until now, we have conducted an overall examination of response differences between survey

modes. From this point forward, we shall examine what form of questions result in particular

response differences and why such differences occur.

（1）Responses to Multiple-answer Questions

Differences exist in the responses to multiple-answer questions when comparing the mail

9

survey with face-to-face interviews. Figure 3 shows the total percentage for each response option

(excluding “nonresponse,” “other,” and “not listed here”) to multiple-answer questions and

calculates the total rate. The mean number of selected response options can be calculated by

dividing this rate by 100.

Figure 3: Total Rate for Multiple-answer Questions Rate

* The total rate for multiple-answer questions is calculated by totaling the percentages for each response option (excluding “Non-response,” “Other,” “Not listed here,” etc.).

* Except for Question 31, a portion of response options differ for Mail A and Mail B. Therefore, the response rate is not calculated for Mail All.

The mail surveys had a higher rate for Question 3 (Television Viewing Genre), Question 12B

(Evaluation of Programming (Commercial broadcasting) and Question 31B (Possession of Media

Equipment). Conversely, Face-to-Face had a higher rate for Question 12A (Evaluation of

Programming (NHK)) and Question 17 (Non-Family Communication). It can be assumed that this

high rate could have been influenced by the concept of social desirability as discussed below, and

by NHK acting as the surveyor.

Depending on the question, the rate to multiple-answer questions in the mail surveys was

higher than Face-to-Face in some cases and lower in other cases. A consistent trend was not

observed.

（2）Social Desirability

Generally speaking, bias can sometimes occur in surveys as the result of social desirability.

This is particularly true for face-to-face interviews. The presence of an interviewer results in a

strongly apparent trend of respondents selecting the socially desirable response.

For example, respondents who abstained from voting may answer that they voted, or smokers

may deny the fact that they smoke. Such responses which are easily accepted by society tend to be

438

259

183

191

512

485

463

242

196

162

506

501

458

237

196

170

501

487

504

494

Question 3: Television Viewing Genre

Question 12A: Evaluation of Programming (NHK）

Question 12B: Evaluation of Programming (Commercial

broadcasting)

Question 17: Non-Family Communication

Question 31A: Possession of Media Equipment

Question 31B: Possession of Media Equipment 2

Face-to-Face Mail A Mail B Mail All

%

10

selected (Groves et al., 2004).

The influence of social desirability could be observed in the 2008 NHK Experimental Survey.

For example, for the item regarding “Essential Communication” in the face-to-face interview, the

number of respondents selecting “Speaking with Friends” was greater than for the mail mode.

The 2010 experiment also showed that, compared to the Mail All, there were questions in

Face-to-Face which appear to have been influenced by social desirability. Examples of such

questions are listed below.

・Family Communication (Question 16): Many respondents answered that they often speak with

family members.

・Non-Family Communication (Question 17): Many respondents answered that they often speak

with people other than family members.

・Degree of Satisfaction with Society (Question 27): Many respondents indicated a high degree of

satisfaction with society.

・Degree of Satisfaction in Life (Question 28): Many respondents indicated a high degree of

satisfaction in life.

To use “Degree of Satisfaction in Life” as an example, many respondents in Face-to-Face

selected “Completely satisfied.” The number of respondents selecting “Somewhat dissatisfied” and

“Completely dissatisfied” was smaller. (Figure 4)

Figure 4: Degree of Satisfaction in Life

The influence of social desirability can also be seen in Question 6D (“I make jokes to the

television while watching”), which asks the modern way of watching television. The percentage of

respondents selecting either “Often” or “Sometimes” was 35 percent for Face-to-Face, 48 percent

for Mail A and 57 percent for Mail B. As shown, the percentage is higher for the Mail All than for

Face-to-Face. During interviews, the presence of an interviewer may have made respondents feel

embarrassed to answer that they make jokes to the television. When examining the percentage by

age groups, the young age group and middle age group for Mail B showed a difference of 20

percent or more when compared to Face-to-Face (Figure 5).

16

10

60

58

18

27

3

5

3

1

Face-to-Face

Mail All

Completely satisfied Somewhat satisfied

Somewhat dissatisfied Completely dissatisfiedDon’t know, nonresponse

%

11

Figure 5: Modern Television Viewing: Assimilation (Often + Sometimes / by age group)

（3）Questions Related to the Surveyor

In the case of questions related to the surveyor, there is a trend in which responses which seem

to fit the expectations of the surveyor are selected. This is also related to social desirability. This

trend is particularly strong during face-to-face interviews because responses are given directly to

the interviewer of the surveyor. When comparing Face-to-Face with the Mail All, the following

response trends occurred for questions related to evaluation/images of NHK and television.

・Image of Television Influence (Question 11): Many respondents indicated a positive image of

television influence.

・Evaluation of Programming (Question 12A, B): Many respondents indicated a positive impression

of NHK programming. Few respondents indicated a positive impression of commercial

television.

・Necessity of NHK (Question 24A): Many respondents indicated the necessity of NHK.

・Necessity of Commercial Broadcasting (Question 24B): Few respondents indicated the necessity

of commercial broadcasting.

Now, let’s examine a few of the examples listed above. First, regarding the “Necessity of

NHK” (Figure 6), many respondents selected “Definitely necessary” for the Face-to-Face. Only a

few respondents selected “Not necessary very much” or “Not necessary at all.” There is a

possibility that respondents showed consideration towards the NHK interviewer.

Figure 6: Necessity of NHK

46

35 27

62

47 36

69

59

45

0

20

40

60

80

100

Young Middle Upper

Face-to-Face Mail A Mail B

(%)

38

30

47

51

10

16

2

4

3

1

Face-to-Face

Mail All

Definitely necessary Somewhat necessary Not necessary very much

Not necessary at all Don’t know, nonresponse

%

12

Figure 7 shows the percentage of respondents by age group who answered that NHK is

“Definitely necessary” or “Somewhat necessary.” For middle and upper age groups, there was no

difference in the percentage depending on the survey mode. However, in the case of the young age

group, the necessity of NHK is higher for Face-to-Face. Due to the fact that the unit response rate

of the young age group was quite low for the face-to-face interview, this phenomenon might be

related to the possibility that many respondents to the survey had a favorable opinion towards NHK.

Figure 7: Necessity of NHK (Definitely + Somewhat necessary / by age group)

In the case of “Image of Television Influence,” the number of responses regarding positive

images was greater for Face-to-Face. For example, during Face-to-Face, many respondents

answered that television had strengthened their ability to think critically and had caused them to

prioritize spiritual wealth (Figure 8 and Figure 9). Only a few respondents answered that television

had caused people’s lifestyle to become standardized (Figure 10). Similar to the “Necessity of

NHK,” responses may have been influenced by a psychological process in which respondents

attempt to show a favorable attitude towards television to the NHK interviewer.

Figure 8: Image of Television Influence (Critical Thinking Ability)

81 82 90

74 78

87

0

20

40

60

80

100

Young Middle UpperFace-to-Face Mail All

(%)

34

29

14

22

48

49

4

1

Face-to-Face

Mail All

Strengthened Weakened

Neither of the above Don’t know, nonresponse

%

13

Figure 9: Image of Television Influence (Material or Spiritual)

Figure 10: Image of Television Influence (Standardization)

The 1976 and 2008 NHK Experimental Surveys report that “during self-administered surveys,

respondents avoid extreme response options and moderate response options are selected.” However,

this experiment did not show such trends. One reason for this is that the current survey was mainly

composed of questions related to television, thus limiting emotional resistance to selecting extreme

response options.

5. Existence/Non-existence of Middle Options

（1）Comparison of Middle Options for Mail A and Face-to-Face

Until now, we have examined the response difference caused by differences in the survey

mode, i.e., mail and face-to-face interview. From here on, we shall consider the difference in

response distributions of survey modes as caused by the existence/non-existence of middle options.

First, in order to compare the middle option rates in self-administered surveys and

interviewer-administered surveys, we will compare Face-to-Face with Mail A, which provided

middle options. Before comparing the two survey modes, Table 7 summarizes the presentation form

of response options and the middle option rate for questions which provided middle options in Mail

A. The presentation form of response options in Face-to-Face can be divided into 4 main types: a)

cases in which the respondent is not provided with showcards (12 items), b) cases in which

showcards exist only for questions (17 items), c) cases in which showcards exist for both the

question and response options, but middle options including “Neither yes nor no” are not provided

(4 items), and d) cases in which showcards exist for both the question and response options

including “Neither yes nor no” (5 items).

25

17

26

32

44

49

5

1

Face-to-Face

Mail All

Spiritual wealth Material wealth


%

12

10

37

48

45

40

6

2

Face-to-Face

Mail All

Individualized Standardized


%

14

Table 7: Presentation Form of Response Options and Middle option rate in Mail A and Face-to-Face

* The Survey (Face-to-Face Interview) included a) no showcard(s), b) showcard(s) for questions only, c) showcard(s) for both questions and response options (however, “Neither yes nor no” was not provided as a response), and d) “Neither yes nor no” provided as a response.

Presentation

of showcards*Presentation form of response

optionMail A

%Face-to-Face %

Mail A –Face-to-Face %

A: Do you watch alone?Often watch alone/ Often watchwith others

17.4 11.4 6.0

B: Do you watch at a specific time?Usually at a specific time/timevaries

4.4 2.0 2.4

C: Do you select programs yourself?Often select myself/ Often selectedby others

18.1 15.8 2.3

D:Do you manage your time forwatching television?

Often/Almost never 21.7 10.4 11.3

E: Do you become engrossed? Often/Almost never 31.4 15.4 16.0

F:Do you only watch programs whichinterest you?

Only watch programs which I aminterested in/ Also watch programswhich I am not interested in

16.6 10.4 6.2

G: Do you want to watch alone? Want to watch alone/with others 38.9 31.2 7.7

H:Do you feel uneasy unless thetelevision is on?

Feel uneasy/ Don’t feel uneasy 11.2 5.6 5.6

7 Often/Almost never 19.4 10.4 9.0

16 Agree/Disagree 21.2 12.8 8.4

A: Selection Yes/No 28.6 15.1 13.5

B: Instant Satisfaction Yes/No 22.5 12.5 10.0

A: Daily Information Agree/Disagree 19.5 9.5 10.0

B: Routine and Rhythms Agree/Disagree 23.8 12.7 11.1

C: Example for Ways of Living Agree/Disagree 38.6 19.9 18.7

D: Center of Family Communication Agree/Disagree 25.5 12.2 13.3

E: Conversational Topics Agree/Disagree 19.0 8.0 11.0

F: Comprehension of Opinions Agree/Disagree 24.8 12.0 12.8

G: Alternative Experience Agree/Disagree 27.6 13.6 14.0

H: Conversational Partner Agree/Disagree 18.1 9.1 9.0

A: ValidityInformation which I am happy toknow/ Information which is notparticularly useful

19.2 6.7 12.5

B: Excess

Information is useful when makingdecisions/ There is an excess ofinformation which makes meconfused

16.4 6.8 9.6

C: ValueI don’t mind spending money toacquire useful information/ I don’twant to spend money

20.4 6.9 13.5

A: Convey Facts? Agree/Disagree 37.8 17.6 20.2

B:Inducement through MassCommunication

Agree/Disagree 17.1 9.1 8.0

TelevisionCharacteristics

C: Realism Agree/Disagree 31.0 18.7 12.3

A: Work or Leisure?Work (housework) or studying./Hobbies or leisure

24.0 9.4 14.6

B: Personal or Society? Personal lifestyle/Society 17.1 7.0 10.1

C: Future or Present? Future/Present 18.4 7.2 11.2

A: Environment Often/Sometimes/Almost never 6.1 1.7 4.4

B: Glimpse Often/Sometimes/Almost never 8.4 2.1 6.3

C: Reference Often/Sometimes/Almost never 9.2 2.5 6.7

D: Assimilation Often/Sometimes/Almost never 4.3 1.2 3.1

A: Critical Thinking Ability Strengthened/Weakened 45.5 47.5 －2.0

B: Standardization Individualized/Standardized 39.7 45.2 －5.5

C: Violence Curbed/Aggravated 53.5 51.9 1.6

D: Material or Spiritual Spiritual wealth/Material wealth 48.3 43.8 4.5

E: Interest in Politics and Society Strengthened/Weakened 24.1 21.4 2.7

Question no. / Question title

a

5TelevisionViewing Habits

Channel Surfing

Family Communication

26Degree ofResponse

b

9Benefits ofTelevision

22RegardingInformatization

23

MassCommunication

25Orientation ofInterest

c 6Modern TelevisionViewing

d 11Image ofTelevisionInfluence

15

Except for d, middle options are not provided through showcards in Face-to-Face. After the

interviewer elicited a response from the respondent, the response was classified as “Neither yes nor

no” when judged as appropriate.

In the following section, we shall examine the middle option rate for Face-to-Face and Mail A,

taking the previously mentioned variations of presentation form into consideration.

Comparison of Middle option rates

In order to ascertain an overview of middle option rates in self-administered surveys and

interviewer-administered surveys, we used a correlation coefficient and scatter diagram (Figure 11)

to check the extent that the middle option rate agreed with Face-to-Face. The subject of this

analysis was 38 questions in which the middle option “Neither yes nor no” was provided for Mail A.

Figure 11: Comparison of Middle Options in Face-to-Face and Mail A

A correlation coefficient of 0.907 was confirmed for the middle options of Mail A and

Face-to-Face. The degree of similarity was slightly low when compared with the 0.976 correlation

for responses of Mail A and Face-to-Face (see p.5). The scatter diagram shows that the middle

option rate was higher for Mail A, which provided a “Neither yes nor no” response.

A similar tendency can be observed when comparing the distribution of middle options for

Mail A and Face-to-Face (Table 8). An middle option rate that is less than 20 percent occurs in 47

percent of cases for Mail A and 84 percent for Face-to-Face. In other words, Face-to-Face has a

lower middle option rate. This is a valid result when considering that, excluding d, Face-to-Face is

composed of questions which do not provide middle options.

Mail A

Face-to-Face

(%)

16

Table 8: Distribution of Middle Options in Mail A and Face-to-Face

Middle Options(%)

Mail A Face-to-Face No. of

response options

% Cumulative

%


% Cumulative

%

Less than 10 5 13.2 13.2 16 42.1 42.1 10.0－19.9 13 34.2 47.4 16 42.1 84.2 20.0－29.9 11 28.9 76.3 1 2.6 86.8 30.0－39.9 6 15.8 92.1 1 2.6 89.5 40.0－49.9 2 5.3 97.4 3 7.9 97.4

50.0 or greater 1 2.6 100.0 1 2.6 100.0 Total 38 100.0 38 100.0

Questions with a High Middle option rate

First, we will specifically examine questions with a high middle option rate “Neither yes nor

no” regardless of the survey mode (Table 7).

Question 11 (Image of Television Influence) is one question that had a high rate for “Neither

yes nor no” in both the Mail A and Face-to-Face. In the case of Face-to-Face, the rate is

conspicuously high because respondents are provided with showcards that includes “Neither yes nor

no.” However, even in the case of Mail A, Question 11 has a high rate when compared to other

questions. This question asks about the image of television influence including whether television

curbs or aggravates delinquency and violence among juveniles. It seems that this question was

difficult for respondents to answer.

Another question for which middle options were selected regardless of the survey mode was

Question 5G, which asks whether or not respondents “want to watch television alone”. In the case of

Question 5G, the large number of selected middle options is probably due to the fact that

respondents want to watch television alone or want to watch together with others depending on the

situation. As already stated, for the face-to-face interview, after the interviewer has elicited a

response from the respondent, the response was classified as “Neither yes nor no” when judged as

appropriate. Therefore, it can be inferred that many respondents answered that their desire to watch

television alone depends on the situation.

Question 23 also had a relatively large number of responses for middle options. Respondents

were asked questions such as “Is the information conveyed by mass communication almost entirely

based on facts?” and “Does watching television give you a sense of realism like you are actually

viewing or hearing the contents?” It can be inferred that this was also a difficult question to answer.

Questions with a Low Middle option rate

Conversely, Question 6 had a low middle option rate in the case of both the Mail A and

Face-to-Face. This question asks respondents about the modern way of watching television.

In Face-to-Face, a 3-point scale with the responses “Often”, “Sometimes” and “Almost never”

was provided for Question 6. It can be inferred that it was easy for respondents to select one of the

prepared response options.

17

Additionally, this question was probably easy to answer because it asked about usual behavior

related to television viewing. For example, respondents were asked to reply to statements such as “I

turn on the television as soon as I return home” and “I turn on the television and watch only when

something interests me.”

Even in the case of Mail A, the middle option rate of Question 6 was conspicuously low when

compared to other questions. When asking about frequency, it may not be necessary to include the

middle option “Neither yes nor no”. However, this may be limited only to questions related to

actual conditions.

For Face-to-Face, other than Question 6, questions with relatively few responses to middle

options were Question 5B (Television Viewing Habits: Do you watch at a specific time?), Question

22 (Regarding Informatization), and Question 25 (Orientation of Interest). These questions were

most likely easy for respondents to answer for the following reasons: Question 5B asks about actual

conditions regarding the time at which television is viewed. Question 22 provides Option 1 and

Option 2, and regardless which response option is selected the contents are innocuous. Question 25

concerns a topic which is vaguely considered during daily life. According to Yamada (2010), “If the

assumption that ‘almost all respondents have thought about the problem and are capable of

producing a response’ is valid, it is preferable not to include the response option ‘Neither yes nor

no.’” Although the middle option rate is fairly high in Question 25 of the Mail A, there are cases in

which it is preferable not to provide middle options for problems which are considered by

respondents during daily life.

Cases with Difference in the Middle option rate

In Mail A and Face-to-Face, questions with particularly large differences in rates are Question

23A, which asks whether the information conveyed by mass communication is almost entirely

based on facts, and Question 9C, which asks whether television provides examples for ways of

living and behavior. The response difference for Question 23A is 20 percent and Question 9C is 19

percent. Mail A has a higher middle option rate. The rate is 38–39 percent for both questions in

Mail A, and it can be said that respondents were unsure of how to answer. Conversely, in

Face-to-Face, only questions were provided on the showcards. Respondents were instructed to

select one of two options “I agree” or “I disagree” when answering the questions. As a result, the

middle option rate was much less than the mail survey and a difference occurred between the survey

modes.

（2）Influence Exerted on Other Response Options by Middle Options

Response Difference Depending on the Existence/Non-existence of Middle Options

Next, in order to consider the influence exerted on other response options by the

existence/non-existence of middle options, analysis was performed for 38 questions. The questions

18

consisted of those for which “Neither yes nor no” was provided as a response in Mail A. A

comparison was made between Mail A and Mail B, which did not provide middle options. (Middle

options and nonresponses were excluded from the comparison.)

Table 9 shows the distribution of absolute values for response differences. Response

differences range from less than 5 percent to 20 percent or more. Furthermore, when compared to

Mail B which does not provide middle options, the responses for Mail A were 10 percent lower on

average. The difference between the minimum and maximum value, as well as the standard

deviation regarding the comparison of Mail A and B was large (Table 10).

Table 9: Distribution of Absolute Value for Response Differences in Mail A and Mail B


difference

Mail A – Mail B No. of

response options

% Cumulative

%

Less than 5.0 17 24.3 24.3 5.0-9.9 19 27.1 51.4

10.0-14.9 18 25.7 77.1 15.0-19.9 12 17.1 94.3

20.0 or greater 4 5.7 100.0 Total 70 100.0

*Only questions in Mail A for which “Neither yes nor no” was provided as a response option (excluding nonresponses and middle options)

Table 10: Minimum, Maximum, Average and Standard Deviation for Response Differences in Mail A

and Mail B Mail A – Mail B

Minimum －22.2 Maximum 3.7 Average －9.8

Standard deviation 6.3 * Only questions in Mail A for which “Neither yes nor no” was provided as a response option (excluding nonresponses and middle options)

When examining a breakdown for questions with small response differences (Table 11), there

are many questions which ask about actual conditions. Examples are “Do you usually watch

television at a specific time?” (Question 5B), “I turn on the television as soon as I return home”

(Question 6A), and “I turn on the television and watch only when something interests me”

(Question 6B).

19

Table 11: Questions with Small Difference in Mail A and Mail B Question

no. Question Response option

Mail A – Mail B %

５B Television Viewing Habits

B: Do you watch at a specific time?

1. Usually at a specific time 0.2

22B Regarding Informatization

B: Excess 2. Option 2: There is an excess of

various information which often makes me confused

0.3

６D Modern Television Viewing

D: Assimilation 1. Often 0.6

６B Modern Television Viewing

B: Glimpse 3. Almost never 1.0

６A Modern Television Viewing

A: Environment 3. Almost never 1.2

６A Modern Television Viewing

A: Environment 1. Often 2.2

６B Modern Television Viewing

B: Glimpse 1. Often 2.4

Conversely, questions with a large response difference (Table 12) include “Is the information

conveyed by mass communication almost entirely based on facts?” (Question 23A), “Does

television provide examples for ways of living and behavior?” (Question 9C), and “Do you want to

watch television alone?” (Question 5G). In general, regarding these questions, Mail A had a high

middle option rate, and thus, respondents might have been unsure how to respond.

Table 12: Questions with Big Difference in Mail A and Mail B

Question no.

Question Response option Mail A –

Mail B %23A Mass Communication A: Convey facts? 2. Disagree 22.2

９C Benefits of Television C: Example for Ways of

Living 2. Disagree 20.8

26A Degree of Response A: Selection 2. No 20.4

９F Benefits of Television F: Comprehension of

Opinions 1. Agree 20.4

５G Television Viewing Habits

G: Do you want to watch television alone?

1. Want to watch alone 19.4

９C Benefits of Television C: Example for Ways of

Living 1. Agree 19.0

Overall Response Trends and Existence/Non-existence of Middle Options

Until now, we have confirmed the distribution and magnitude of response differences for Mail

A and Mail B. We will now examine specific response trends.

Overall, essentially the same response trends as Mail A can be observed in Mail B, which does

not provide middle options. However, when middle options are not provided, the ‘pro’ and ‘anti’

sides of the responses result in a more emphasized way. This result supports the previous research

of Kalton et al (Suzuki 2000) which states the following: “Providing middle options results in an

increase of responses for middle options. However, this increase comes proportionally from both of

the extreme opposite responses. Therefore, there is no change in the balance between ‘pro’ and

20

‘anti.’” In other words, although the provision (non-provision) of middle options causes a decrease

(increase) in the rate for other response options, there is no change in the balance between the

extreme opposite response options.

There was no clear trend that particular response options were selected as a result of

existence/non-existence of middle options. However, like in the case of Question 6, the response

option “Sometimes” attracted many respondents in Mail B. Although Mail B does not provide

middle options, it does provide a 3-point scale of response options: “Often,” “Sometimes,” and

“Almost never.” In this case, “Sometimes” is in the midpoint of the 3 points, and there is a

possibility that it took the position of a middle option.

Middle option rate Affected by Order Effects of Questions

The order and layout of questions in the questionnaire result in an order effect of questions.

This order effect may influence the middle option rate. In Mail A, middle options are provided for

Questions 5 to 9 (excluding Question 8). In contrast, such responses are not provided in Mail B.

Therefore, the first middle option provided in Mail B is the “Neither yes nor no” response for

Question 11. As a result, this middle option in Mail B attracted more respondents than in Mail A

(Figure 12). According to a report by the Institute of Statistical Mathematics, when a question

which provides a middle option is placed between questions without middle options, there is the

possibility that respondents feel that the middle option is appealing (Maeda, 2005). There is a

possibility that this kind of psychological process was at work in the current experiment as well.

Figure 12: Image of Television Influence (Critical Thinking Ability)

6. Summary

Response Rate

In the current experimental survey, for young age groups of both men and women, the

response rate for the mail survey was higher than for the face-to-face interview. These results

support the concept that the mail survey is a more effective method for young age groups in which

face-to-face interviews often do not lead to valid responses.

30

28

23

20

46

52

2

1

Mail A

Mail B

Strengthened Weakened


%

21

Comparison of Mail Surveys and Face-to-face Interviews

The following types of response trends were observed when comparing the mail survey and

face-to-face interviews.

・Overall, there were fewer nonresponses in the mail survey.

・Socially desirable responses tend to be selected in face-to-face interviews.

・For questions related to the surveyor, responses which seem to meet the expectations of the

surveyor tend to be selected in face-to-face interviews.

In the current experimental survey, for face-to-face interviews in which the interviewer asks

questions directly to the respondent, there was a trend in which responses which seemed to meet the

expectations of the surveyor (NHK) were selected more often than for the mail survey. NHK has

used a variety of survey modes to conduct numerous surveys related to NHK. However, careful

interpretation is necessary when referring to these kinds of survey results for content related to the

surveyor.

Existence/Non-existence of Middle Options

The current experimental survey attempted to verify whether or not differences occur in

overall response trends depending on the existence/non-existence of middle options in a

self-administered survey. The following differences were observed from survey results.

・When middle options not provided for the face-to-face interview are provided for the mail survey,

the middle option rate tends to be higher than the face-to-face interview.

・The middle option rate tends to be lower for questions which concern actual conditions or for

questions which are considered by the respondent on a daily basis.

・When there is a succession of questions without middle options, the order effect of subsequent

questions may influence the middle option rate.

・When middle options are not provided, there may be a decrease in the number of nonresponses.

・Even for the same self-administered survey, the existence/non-existence of middle options may

result in significant differences for other selected response options.

・Although the provision (non-provision) of middle options causes a decrease (increase) in the rate

for other response options, there is no change in the balance between the extreme opposite

response options.

・Depending on the existence/non-existence of middle options, there may be questions for which

certain response options attract many respondents. However, it is difficult to find a principle for

this phenomenon from the results of the experimental survey.

Furthermore, there was a high middle option rate when they were provided in the list of

response options for face-to-face interviews. Moreover, even if such response options are not

provided, the middle option rate may increase due to the contents of questions and the form in

which response options are presented.

The results of the current survey do not enable us to make a blanket statement regarding in

22

what cases middle options should be provided or not provided. However, when considering whether

or not to provide middle options in designing a questionnaire, it is necessary to consider a variety of

factors, not only question content, but also order effect, and the number/form of response options.

(Hiroko Murata/Noriko Onodera)

Notes: 1) The following members of the Public Opinion Research Division were involved in planning, implementation and

analysis for the 2010 experimental mail survey: Hiroshi Aramaki, Noriko Onodera, Kei Kono, Hiroshi Kojima, Hiroshi Nakaaki, Kumiko Nishi, Miwako Hara, Hiroko Murata.

2) Detailed results for the Comparative Experimental Survey 2008 were reported in The Annual Bulletin of NHK Broadcasting Culture Research Institute 2010 (Survey Methodology Project 2010).

3) The Japanese and Television 2010 Survey consisted of two surveys using a face-to-face interview and a pick-up/drop-off survey. Questions were different for each survey. The objective of the experimental survey was to compare the self-administered mode (mail survey) and the interviewer administered mode (face-to-face interview). Therefore, the same questions were used as in the face-to-face interview.

4) Mail All: χ2=21.79, df=5, p<.001, Mail A: χ2=10.44, df=5, n.s, Mail B: χ2=12.47, df=5, p<.05, Face-to-Face: χ2=30.72, df=5, p<.001.

5) For each single answer question of the Mail All, respondents are instructed to “Circle one answer.” However, in the case of questions 19A, B, C, D, E, F and G, the unit of the instructions was not appropriate. As a result, a significant number of respondents answered the question under the misimpression that multiple answers were allowed.

References ・Survey Methodology Project “Effects of Survey Mode on Survey Administrations and Responses-Comparison of Face-to-Face

Interview, Drop-off/Pick-up and Mail Surveys-” The Annual Bulletin of NHK Broadcasting Culture Research Institute 54, 2010

・Groves, R.M., Fowler, F.J., Couper, M .P., Lepkowski, J.M., Singer. E., Tourangeau, R., Survey Methodology, Wiley, 2004 ・Maeda, Tadahiko, “A Study on the Characteristics of Mail Survey: A Comparison with Face-to-Face Interviewing”,

Proceedings of the Institute of Statistical Mathematics , Vol. 53, No. 1, 2005 ・NHK Public Opinion Research Institute “Comparative Study on Survey Methodology”, 1977 ・Suzuki, Tatsuzo, “The Effect of the Question on Survey Responses: A Review—Kalton and Schuman (1)” Shin Joho,

83, 2000 ・Yamada, Kazunari, “Kikikata no gijutsu: Risaachi no tame no chosahyo sakusei gaido”(in Japanese) [The Skill of

Asking: Guidelines for Creating Questionnaires for Research], Nikkei Publishing Inc., 2010

23

The Japanese and Television 2010 Experimental Survey: Questions & Marginal Count Results

Mail All Mail A Mail B Respondents allocated to

questions which provide middle options

Respondents allocated to questions which do not provide middle options

Survey Period March 3rd (Wed.) to April 16th (Fri.), 2010

Survey Method Mail

Population Japanese aged 16 and olderSample 1,800 900 900

(12 respondents×150 units) (6 respondents×150 units) (6 respondents×150 units)

No. of Respondents (Response Rate)

1,215 respondents (67.5%)

609 respondents(67.7%)

606 respondents (67.3%)

<Regarding Notation> ・The questions listed were used for the mail survey. ・Refer to the August issue of The NHK Monthly Report on Broadcast Research for the marginal count of the

face-to-face interview. ・ Response options shown in the showcard for the face-to-face interview are printed in boldface. ・ Numbers shown in the table are the number of responses for each response option, divided by the total number of

valid responses for each survey, displayed as a percentage (%) (rounded to the second decimal place). ・ Inequality signs in the table are statistical test results from a comparison of Mail A and Mail B (95% confidence

level). ・When the response options differ for Mail A and Mail B, values for the Mail All are not displayed. ・A result of “-” shows that there were no response options. ・ Although Question 19 was a single answer question, the design of the mail mode survey caused many respondents to

misunderstand that it was a multiple answer question. Therefore, Question 19 is omitted from the analysis. -Number of Hours Spent Watching TV- Question 1: Excluding days off, how many hours do you watch television per day on a weekday? (Circle one)

Mail All Mail A Mail B

1. About 1 hour 9.3 8.2 10.4 %

2. About 2 hour 23.2 24.6 21.8 3. About 3 hour 19.5 18.9 20.1 4. About 4 hour 17.3 17.6 17.0 5. About 5 hour 10.9 10.7 11.1 6. About 6 hour 7.0 7.7 6.3 7. About 7 hour 2.1 2.3 2.0 8. About 8 hour 3.0 3.0 3.0 9. More than 9 hours 3.5 3.4 3.6

10. None / Almost none 3.5 3.0 4.1 11. No response 0.7 0.7 0.7

-NHK or Commercial Broadcasting- Question 2: Which do you watch more, NHK programs or commercial broadcasting programs? (Circle one)


1. I watch NHK much more 9.1 8.2 9.9 %

2. I watch NHK slightly more 9.3 9.2 9.4 3. I watch about the same amount of NHK and

commercial broadcasting 20.7 20.2 21.3

4. I watch commercial broadcasting slightly more 32.8 34.5 31.0 5. I watch commercial broadcasting much more 27.0 26.8 27.2 6. No response 1.2 1.1 1.2

24

-Television Viewing Genre (multiple answer)- Question 3: What television programs do you normally watch? (Circle one or more)

Mail A Mail B

1. News, news shows 75.9 77.2 %

2. Weather forecasts 61.1 62.2 3. Programs on politics, economics and society 26.4 31.2 4. Dramas 51.7 ＞ 45.5 5. Theatrical movies 22.3 19.6 6. Quiz and game shows 25.8 23.9 7. Vaudeville shows including rakugo and manzai 11.3 10.1 8. Entertainment, comedy and variety shows 35.6 35.1 9. Programs featuring songs and music 29.2 31.4

10. Animation and comic 13.5 12.2 11. Sports programs 31.9 30.5 12. Educational programs on foreign languages,

skills/crafts, hobbies, etc. 7.7 8.9

13. Cultural programs on nature, history, travel, science, etc.

24.5 27.4

14. Lifestyle and practical information programs 17.4 16.3 15. Morning and daytime tabloid shows 28.9 26.7 16. Other (please specify) 2.1 ― 17. No response 1.1 1.3

-Commercial Viewing- Question 4: How do you watch television commercials? Which of the following responses best matches your attitude?

(Circle one) Mail All Mail A Mail B

1. Enjoy watching 20.3 20.2 20.5 %

2. Watch with indifference 50.9 51.7 50.0 3. Watch but dislike 16.6 16.6 16.7 4. Almost never watch 10.8 10.0 11.6 5. No response 1.4 1.5 1.3

-Television Viewing Habits- A: Do you watch alone? Question 5A: Please answer the following series of questions on your television viewing habits. (Circle one response for each question from A to H) A: Do you often watch television alone? Or do you often watch with other people?

Mail A Mail B

1. Often watch alone 43.2 ＜ 50.3 %

2. Often watch with others 38.6 ＜ 48.8 3. Neither of the above 17.4 ― 4. No response 0.8 0.8

B: Do you watch at a specific time? Question 5B: B. Do you usually watch television at a specific time? Or does the time when you watch television vary?

Mail A Mail B

1. Usually at a specific time 80.8 81.0 %

2. Time varies 13.8 ＜ 18.0

3. Neither of the above 4.4 ―

4. No response 1.0 1.0

25

C: Do you select programs yourself? Question 5C: C. Do you often personally select the programs that you view? Or do other people often select the programs?

Mail A Mail B

1. Often select myself 66.2 ＜ 79.7 %

2. Often selected by others 15.1 19.3 3. Neither of the above 18.1 ― 4. No response 0.7 1.0

D: Do you manage your time for watching television? Question 5D: D. Do you often manage your time so that you are able to watch television? Or do you almost never manage your viewing time?

Mail A Mail B

1. Often 24.0 27.6 %

2. Almost never 53.5 ＜ 71.1 3. Neither of the above 21.7 ― 4. No response 0.8 1.3

E: Do you become engrossed? Question 5E: E. Do you often become engrossed when viewing television? Or do you almost never become engrossed?

Mail A Mail B

1. Often 39.7 ＜ 55.4 %

2. Almost never 28.1 ＜ 43.1 3. Neither of the above 31.4 ― 4. No response 0.8 1.5

F: Do you only watch programs which interest you? Question 5F: F. Do you only watch television programs which interest you? Or do you watch programs which are not of particular interest?

Mail A Mail B

1. Only watch programs which I am interested in 57.1 ＜ 70.3 %

2. Also watch programs which I am not interested in 25.6 28.7 3. Neither of the above 16.6 ― 4. No response 0.7 1.0

G: Do you want to watch alone? Question 5G: G. Do you want to watch television alone? Or do you want to watch with others?

Mail A Mail B

1. Want to watch alone 37.4 ＜ 56.8 %

2. Want to watch with others 22.7 ＜ 41.4 3. Neither of the above 38.9 ― 4. No response 1.0 1.8

H: Do you feel uneasy unless the television is on? Question 5H: H. Do you feel uneasy unless the television is on? Or do you not experience such feelings of uneasiness?

Mail A Mail B

1. Feel uneasy 19.2 ＜ 24.1 %

2. Don’t feel uneasy 68.1 ＜ 74.9 3. Neither of the above 11.2 ― 4. No response 1.5 1.0

26

-Modern Television Viewing- A: Environment Question 6A: Do your television viewing habits coincide with the behaviors listed below? Which of the following apply? (Circle one response for each question from A to D) A. I turn on the television as soon as I return home

Mail A Mail B

1. Often 39.4 41.6 %

2. Sometimes 26.9 ＜ 32.2 3. Almost never 26.3 25.1 4. Neither of the above 6.1 ― 5. No response 1.3 1.2

B: Glimpse Question 6B: B. I turn on the television and watch only when something interests me.

Mail A Mail B

1. Often 26.8 29.2 %


C: Reference Question 6C: C. I watch while comparing with other programs and information regarding the cast.

Mail A Mail B

1. Often 14.4 ＜ 19.5 %


D: Assimilation Question 6D: D. I make jokes to the television while watching

Mail A Mail B

1. Often 12.0 11.4 %


-Channel Surfing- Question 7: Do you often channel surf using the remote control? Or do you almost never engage in channel surf? (Circle one)

Mail A Mail B

1. Often 37.8 ＜ 43.7 %

2. Almost never 40.1 ＜ 52.8 3. Neither of the above 19.4 ― 4. Do not own a remote control 1.1 ＜ 3.0 5. No response 1.6 0.5

27

-Method of Selecting Programs- Question 8: How do you normally select programs when watching television? Please select the option which you think applies most often. (Circle one)


1. Select by channel surfing with remote control 20.9 21.5 20.3 %

2. Select from a list of television programs in newspapers

45.6

46.6 44.6

3. Select from a list of television programs in magazines

1.6

1.8 1.5

4. Select from a list of television programs on the internet

2.1

1.6 2.6

5. I frequently watch the same programs 22.8 21.3 24.3 6. I watch programs selected by family members 5.7 5.3 6.1 7. No response 1.2 1.8 0.7

-Benefits of Television- A: Daily Information Question 9A: This section contains questions on the relationship between television and your lifestyle. Please answer items A to H listed below. (Circle one response for each question from A to H) The Existence of Television A. makes it possible to obtain essential information for daily life.

Mail A Mail B

1. Agree 68.6 ＜ 76.1 %

2. Disagree 10.2 ＜ 23.4 3. Neither of the above 19.5 ― 4. No response 1.6 0.5

B：Routine and Rhythms Question 9B: B. provides a source of routine or rhythm in daily activities.

Mail A Mail B

1. Agree 48.8 ＜ 59.1 %

2. Disagree 25.1 ＜ 40.1 3. Neither of the above 23.8 ― 4. No response 2.3 ＞ 0.8

C：Example for Ways of Living Question 9C: C. provides examples for ways of living and behavior

Mail A Mail B

1. Agree 33.3 ＜ 52.3 %


D：Center of Family Communication Question 9D: D. is useful for family communication.

Mail A Mail B

1. Agree 60.6 ＜ 73.6 %


28

E：Conversational Topics Question 9E: E. makes it possible to obtain conversational topics.

Mail A Mail B

1. Agree 71.1 ＜ 83.7 %


F：Comprehension of Opinions Question 9F: F. makes it possible to comprehend a variety of opinions and ways of thinking.

Mail A Mail B

1. Agree 65.2 ＜ 85.6 %

2. Disagree 8.0 ＜ 13.7 3. Neither of the above 24.8 ― 4. No response 2.0 ＞ 0.7

G：Alternative Experience Question 9G: G. makes it possible to get a taste of experiences not obtainable in reality.

Mail A Mail B

1. Agree 51.2 ＜ 63.2 %


H：Conversational Partner Question 9H: H. is a substitute for a conversational partner.

Mail A Mail B

1. Agree 18.1 ＜ 25.7 %


-Importance of Television- Question 10: Put simply, what meaning does television have for you? Select the response closest to your feelings from those listed below. (Circle one)


1. Essential 46.9 47.5 46.4 %

2. Nothing more than a convenience 46.6 46.8 46.4 3. Nothing special 5.3 3.9 ＜ 6.8 4. A nuisance 0.3 0.3 0.3 5. No response 0.8 1.5 ＞ 0.2

-Image of Television Influence- A: Critical Thinking Ability Question 11A: This section contains questions on the influence of television. Select the response which is closest to your opinion from A to E listed below. (Circle one response for each question A to E) A. Television has ________ people’s ability to think critically.


1. Strengthened 28.7 29.9 27.6 %

2. Weakened 21.6 22.8 20.3 3. Neither of the above 48.6 45.5 ＜ 51.7 4. No response 1.2 1.8 ＞ 0.5

29

B：Standardization Question 11B: B. Television has ___________ people’s lifestyles.


1. Individualized 10.0 9.5 10.4 %

2. Standardized 48.3 48.3 48.3 3. Neither of the above 40.2 39.7 40.8 4. No response 1.5 2.5 ＞ 0.5

C：Violence Question 11C: C. Television has _________ delinquency and violence among youth.


1. Curbed 6.4 9.2 ＞ 3.6 %

2. Aggravated 36.4 35.3 37.5 3. Neither of the above 56.0 53.5 58.6 4. No response 1.2 2.0 ＞ 0.3

D：Material or Spiritual Question 11D: D. Television made people prioritize _____________.


1. Spiritual wealth 17.1 16.3 18.0 %

2. Material wealth 32.3 33.7 30.9 3. Neither of the above 49.4 48.3 50.5 4. No response 1.2 1.8 0.7

E：Interest in Politics and Society Question 11E: E. Television has ___________ people’s interest in politics and society.


1. Strengthened 67.4 68.6 66.2 %

2. Weakened 4.9 5.7 4.1 3. Neither of the above 26.6 24.1 29.0 4. No response 1.1 1.5 0.7

-Evaluation of Programming (NHK) (multiple answer)- Question 12A: A. What is your impression of recent NHK television programming? Circle one or more of the following options which match your impressions. (Circle one or more)

Mail A Mail B

1. Fresh programming 24.5 21.9 %

2. Programming that perfectly matches my feelings 19.5 19.8 3. Entertainment programming is amusing and interesting 15.6 14.2 4. Rapid response to accidents and disasters 59.8 63.4 5. Information programs delve deep into the heart of issues 15.4 14.9 6. Information programs are neutral and fair 26.6 26.1 7. Educational programs are extremely interesting 39.7 37.0 8. Comprehensive coverage of local events and topics 41.1 39.4 9. None of the above 9.2 ―

10. No response 2.0 ＜ 7.1

30

-Evaluation of Programming (Commercial broadcasting) (multiple answer) - Question 12B: B. How about commercial broadcasting television programs? Circle one or more of the following options. (Circle one or more)

Mail A Mail B

1. Fresh programming 23.2 23.9 %

2. Programming that perfectly matches my feelings 28.9 26.9 3. Entertainment programming is amusing and interesting 61.6 63.4 4. Rapid response to accidents and disasters 20.0 19.1 5. Information programs delve deep into the heart of issues 17.7 17.2 6. Information programs are neutral and fair 4.1 3.0 7. Educational programs are extremely interesting 10.5 9.6 8. Comprehensive coverage of local events and topics 29.6 33.3 9. None of the above 11.0 ―

10. No response 3.1 ＜ 6.4

-Degree of Satisfaction with TV- Question 13: Overall, how satisfied are you with current television. (Circle one)


1. Completely satisfied 11.2 12.3 10.1 %

2. Fairly satisfied 63.4 59.9 ＜ 66.8 3. Fairly dissatisfied 21.2 22.3 20.1 4. Completely Dissatisfied 2.9 3.1 2.6 5. No response 1.3 2.3 ＞ 0.3

-Expectations for Broadcasting- Question 14: When classifying television programs into the following 3 categories, which of the following programs would you like to be broadcasted the most? (Circle one)


1. Programs for relaxation and enjoyment 42.3 42.5 42.1 %

2. Programs which cover events and trends in the world

41.0

40.7 41.3

3. Programs useful for acquiring knowledge and education

15.8

15.6 16.0

4. No response 0.9 1.1 0.7

-Amount of Free Time- Question 15: Excluding days off, how many hours of free time do you have at your discretion on a weekday? (Circle one)


1. About 1 hour 8.1 7.6 8.6 %

2. About 2 hour 21.9 21.3 22.4 3. About 3 hour 17.2 17.4 17.0 4. About 4 hour 18.1 20.4 ＞ 15.8 5. About 5 hour 12.2 12.0 12.4 6. About 6 hour 6.9 6.2 7.6 7. About 7 hour 2.1 1.8 2.3 8. About 8 hour 2.8 2.8 2.8 9. More than 9 hours 7.7 7.1 8.3

10. Almost never 2.6 2.8 2.3 11. No response 0.6 0.7 0.5

31

-Family Communication- Question 16: Do you feel that you talk a lot with your family in everyday life? Or do you not feel so? (Circle one)

Mail A Mail B

1. Agree 57.6 ＜ 68.0 %

2. Disagree 16.1 ＜ 25.7 3. Neither of the above 21.2 ― 4. I have no family 4.8 6.1 5. No response 0.3 0.2

-Non-family Communication-(multiple answer) Question 17: From the following options, please select one or more category of people that you talk a lot with in everyday life. (Circle one or more)

Mail A Mail B

1. Relatives 21.3 20.1 %

2. Childhood friends or friends from school 19.7 22.8 3. Neighbors 25.0 25.4 4. Coworkers 43.2 44.7 5. People that you met through work 17.9 20.6 6. People that you met through your family 14.3 15.8 7. Friends involved in same hobbies/activities 20.7 21.0 8. Other (please specify) 7.6 ― 9. I don’t talk with people often 6.4 7.8

10. No response 1.5 2.0 -Frequency of Access to Media- A:Television Question 18A: How much contact do you have with television, newspapers, etc? For items A to J listed below, how often do you watch, listen, read or use each item? (Circle one response for each question from A to J) A .Television (excluding videos, video games, etc.)


1. Every day 88.8 90.1 87.5 %

2. 3 or 4 times per week 4.9 3.9 5.8 3. Once or 2 times per week 2.3 1.6 3.0 4. Once or 2 times per month 0.5 0.5 0.5 5. Almost never or never watch 2.6 2.6 2.6 6. No response 0.9 1.1 0.7

B：Video, DVD, Blu-ray Disc Question 18B: Video, DVD, Blu-ray Disc


1. Every day 7.1 8.7 ＞ 5.4 %

2. 3 or 4 times per week 6.4 5.9 6.9 3. Once or 2 times per week 19.7 21.0 18.3 4. Once or 2 times per month 21.8 21.3 22.3 5. Almost never or never watch 42.8 40.2 45.4 6. No response 2.2 2.8 1.7

C：HDD Question 18C: HDD (Hard Disc Recorder)


1. Every day 8.6 9.7 7.4 %

2. 3 or 4 times per week 5.6 4.8 6.4 3. Once or 2 times per week 11.1 11.2 11.1 4. Once or 2 times per month 6.5 8.0 ＞ 5.0 5. Almost never or never watch 64.9 62.2 ＜ 67.7 6. No response 3.3 4.1 2.5

32

D：Radio Question 18D: D. Radio (including FM and short-wave; excluding when listening to recorded programs)


1. Every day 24.6 24.5 24.8 %

2. 3 or 4 times per week 6.2 7.6 ＞ 4.8 3. Once or 2 times per week 8.2 7.2 9.2 4. Once or 2 times per month 6.9 7.7 6.1 5. Almost never or never listen 53.1 51.7 54.5 6. No response 1.0 1.3 0.7

E：CD, etc. Question 18E: E.CD, MD, cassette tapes


1. Every day 22.9 23.2 22.6 %

2. 3 or 4 times per week 8.7 8.0 9.4 3. Once or 2 times per week 9.3 9.0 9.6 4. Once or 2 times per month 13.1 13.8 12.4 5. Almost never or never listen 44.1 44.2 44.1 6. No response 1.9 1.8 2.0

F：Newspaper Question 18F: F. Newspaper


1. Every day 70.4 70.3 70.5 %

2. 3 or 4 times per week 5.3 5.1 5.4 3. Once or 2 times per week 5.0 5.3 4.8 4. Once or 2 times per month 1.8 1.3 2.3 5. Almost never or never read 17.0 17.6 16.5 6. No response 0.5 0.5 0.5

G：Weekly magazines Question 18G: G.：Weekly magazines (excluding comics and graphic novels)


1. Every day 1.5 1.3 1.7 %

2. 3 or 4 times per week 1.8 2.6 ＞ 1.0 3. Once or 2 times per week 9.4 9.7 9.1 4. Once or 2 times per month 21.0 22.0 20.0 5. Almost never or never read 65.2 63.1 67.3 6. No response 1.2 1.3 1.0

H：Books Question 18H: H.：Books (excluding comics and graphic novels)


1. Every day 12.6 13.8 11.4 %

2. 3 or 4 times per week 8.1 7.4 8.7 3. Once or 2 times per week 10.2 11.2 9.2 4. Once or 2 times per month 26.3 28.6 23.9 5. Almost never or never read 41.7 37.8 ＜ 45.7 6. No response 1.2 1.3 1.0

33

I: Comics Question 18I: I.：Weekly magazines and books featuring comics or graphic novels


1. Every day 3.5 3.6 3.5 %

2. 3 or 4 times per week 4.1 4.3 4.0 3. Once or 2 times per week 7.7 8.2 7.3 4. Once or 2 times per month 15.2 17.1 13.4 5. Almost never or never read 68.2 65.2 ＜ 71.3 6. No response 1.2 1.6 0.7

J：Internet Question 18J: J. Internet (excluding email) (Please include time spent viewing homepages or searching for information on computer or mobile phones.)


1. Every day 32.8 34.5 31.0 %

2. 3 or 4 times per week 7.5 7.6 7.4 3. Once or 2 times per week 9.6 8.0 11.2 4. Once or 2 times per month 6.3 6.1 6.6 5. Almost never or never use 42.5 42.4 42.6 6. No response 1.3 1.5 1.2

-Comparison of Media Utility- Question 19: (omitted) -Essential Media: Number 1- Question 20A: A .What media is essential for you? Consider a variety of factors and select a single response from the media listed below. (Circle one)

Mail A Mail B

1. Radio 3.3 3.0 %

2. CD, MD, cassette tapes 2.1 2.3 3. Television 35.3 33.8 4. Movies, video software 0.2 0.5 5. Newspapers 10.7 9.4 6. Weekly magazines 0.0 0.3 7. Town magazines, information magazines 0.0 0.2 8. Books 3.9 2.6 9. Internet (excluding email) 10.3 11.2

10. Communication with family 28.6 32.0 11. Communication with friends 4.1 4.3 12. Not listed here 0.8 ― 13. No response 0.7 0.3 -Essential media: Number 2- Question 20B: B .Which would you select if you were to choose one more? (Circle one)

Mail A Mail B

1. Radio 3.9 4.5 %

2. CD, MD, cassette tapes 3.0 3.3 3. Television 27.3 29.7 4. Movies, video software 2.0 1.7 5. Newspapers 16.4 18.5 6. Weekly magazines 0.2 0.8 7. Town magazines, information magazines 1.1 0.5 8. Books 6.7 5.8 9. Internet (excluding email) 8.5 7.6

10. Communication with family 11.2 9.9 11. Communication with friends 14.0 14.9 12. Not listed here 2.5 ― 13. No response 3.3 3.0

34

-Comparison of Useful Media- A: Promptness Question 21A: This section contains questions regarding how you obtain news and information. Select a single response from those listed below. (Circle one response for each question from A to D) A .Which form of media conveys information most promptly?

Mail A Mail B

1. Radio 8.7 6.8 %

2. Television 75.0 77.4 3. Newspapers 0.5 1.7 4. Magazines 0.0 0.0 5. Internet (excluding email) 14.0 13.4 6. Conversation with family/friends 1.0 0.5 7. Not listed here 0.0 ― 8. No response 0.8 0.3

B：Ease of Understanding Question 21B: B. Which form of media is the easiest to understand?

Mail A Mail B

1. Radio 2.1 2.1 %


C：Level of Detail Question 21C: C. Which form of media provides the most detailed information?

Mail A Mail B

1. Radio 1.3 1.3 %


D：Selection Question 21D: D. Which form of media makes it easier to select only necessary information?

Mail A Mail B

1. Radio 2.3 1.7 %


35

-Regarding Informatization- A: Validity Question 22A: We are able to acquire a variety of information from sources such as television and newspapers. Items A to C listed below provide positive and negative viewpoints. Which viewpoint is closest to your feelings? (Select one response for each question from A to C) A.

Mail A Mail B

1. Option 1: There is a large amount of information which I am happy to know

66.5 ＜ 77.4 %

2. Option 2: There is a large amount of information which is not particularly useful

13.0 ＜ 20.8

3. Neither of the above 19.2 ― 4. No response 1.3 1.8

B：Excess Question 22B: B.

Mail A Mail B

1. Option 1: A variety of information is useful when making decisions

48.8 ＜ 65.3 %

2. Option 2: There is an excess of various information which often makes me confused

33.2 33.5


C：Value Question 22C: C.

Mail A Mail B

1. Option 1: I don’t mind spending a substantial amount of money in order to acquire useful information

8.7 ＜ 13.5 %

2. Option 2: I don’t want to spend a substantial amount of money on information even if it is useful

68.6 ＜ 85.0


-Mass Communication- A: Convey Facts? Question 23A: What is your opinion regarding the views in A to C listed below? (Circle one response for each question from A to C) A. The information conveyed by mass communication is almost entirely based on facts.

Mail A Mail B

1. Agree 20.4 ＜ 35.6 %


B：Inducement through Mass Communication Question 23B: B. People’s opinions are often swayed in accordance with information conveyed through mass communication without realizing.

Mail A Mail B

1. Agree 74.5 ＜ 83.7 %


36

-Television Characteristics- C:Realism Question 23C: C. Watching television gives me a sense of realism like I am actually viewing or hearing the contents.

Mail A Mail B

1. Agree 30.0 ＜ 45.9 %


-Necessity of NHK- Question 24A: How do you feel about NHK? Pleases select the option which is closest to your feelings from those listed below. (Circle one)


1. Definitely necessary 29.6 29.1 30.2 %

2. Somewhat necessary 50.5 52.5 48.3 3. Not necessary very much 15.6 15.3 16.0 4. Not necessary at all 3.8 2.8 4.8 5. No response 0.5 0.3 0.7

-Necessity of Commercial Broadcasting- Question 24B: B. How do you feel about commercial broadcasting (Circle one)


1. Definitely necessary 28.3 28.4 28.2 %

2. Somewhat necessary 62.0 64.2 59.7 3. Not necessary very much 8.6 6.6 ＜ 10.6 4. Not necessary at all 0.6 0.3 0.8 5. No response 0.6 0.5 0.7

-Orientation of Interest- A: Work or Leisure? Question 25A: On a different subject, when comparing Option 1 and Option 2 as listed in A to C below, which one interests you most? (Select one response for each question from A to C) A.

Mail A Mail B

1. Option 1: Work (housework) or studying 35.8 ＜ 46.9 %

2. Option 2: Hobbies or leisure 38.6 ＜ 51.7 3. Neither of the above 24.0 ― 4. No response 1.6 1.5

B: Personal or Society? Question 25B: B.

Mail A Mail B

1. Option 1: Personal lifestyle 62.9 ＜ 72.6 %

2. Option 2: Society 18.4 ＜ 25.9 3. Neither of the above 17.1 ― 4. No response 1.6 1.5

C：Future or Present? Question 25C: C.

Mail A Mail B

1. Option 1: Future 38.3 43.2 %

2. Option 2: Present 40.6 ＜ 55.3 3. Neither of the above 18.4 ― 4. No response 2.8 1.5

37

-Degree of Response- A:Selection Question 26A: This section contains questions on your personality. Please select one answer for questions A and B. A. When you desire something, do you search until you are satisfied? (Circle one)

Mail A Mail B

1. Yes 44.7 ＜ 52.6 %

2. No 26.3 ＜ 46.7 3. Neither of the above 28.6 ― 4. No response 0.5 0.7

B：Instant Satisfaction Question 26B: B. Are you disturbed when something that you desire cannot be obtained immediately? (Circle one)

Mail A Mail B

1. Yes 18.9 23.3 %

2. No 57.6 ＜ 76.1 3. Neither of the above 22.5 ― 4. No response 1.0 0.7

-Degree of Satisfaction with Society- Question 27: How satisfied are you with society? (Circle one)



2. Somewhat satisfied 35.8 35.8 35.8 3. Somewhat dissatisfied 51.8 51.1 52.5 4. Completely dissatisfied 9.0 9.0 8.9 5. No response 0.9 1.1 0.7

-Degree of Satisfaction in Life- Question 28: Overall, how satisfied are you with your current lifestyle? (Circle one)



2. Somewhat satisfied 57.9 57.8 58.1 3. Somewhat dissatisfied 26.6 27.4 25.7 4. Completely dissatisfied 5.0 3.8 ＜ 6.3 5. No response 0.8 0.8 0.8

-Number of televisions- Question 29: How many televisions do you have in your home? (both color and black & white) (Circle one)


1. 1 24.3 24.8 23.8 %

2. 2 30.0 30.9 29.0 3. 3 24.9 24.0 25.9 4. 4 or more 19.6 19.2 20.0 5. 0 1.0 1.0 1.0 6. No response 0.2 0.2 0.3

38

-No. of Commercial Broadcasting Channels- Question 30: How many channels can you watch on television in your home, including commercial broadcasting channels? (including UHF and satellite broadcasting) (Circle one)


1. No commercial broadcasting, 1 to 2 channels 1.3 1.5 1.2 %

2. 3 channels 3.0 2.8 3.1 3. 4 channels 7.7 7.2 8.3 4. 5 channels 12.2 12.6 11.7 5. 6 channels 10.3 9.7 10.9 6. 7 channels 5.9 5.3 6.6 7. 8 channels 5.6 4.9 6.3 8. 9 channels 2.8 2.3 3.3 9. 10 or more channels 48.7 51.2 46.2

10. No response 2.5 2.5 2.5

-Possession of Media Equipment (multiple answer) Question 31A: Which of the following media equipment do you have in your home? (Circle one or more)


1. Home video recorder 63.2 63.1 63.4 %

2. Television with 25-inch screen or larger 75.1 76.2 74.1 3. Widescreen television 42.8 39.9 ＜ 45.7 4. CATV (television which can receive numerous

channels) 27.0

29.4 24.6

5. Computer/word processor 73.5 72.9 74.1 6. Video game system 46.2 45.8 46.5 7. Television capable of receiving text

broadcasting 20.7

21.3 20.1

8. Television capable of receiving satellite broadcasting

49.6

51.4 47.9

9. Television capable of receiving high-vision broadcasting

39.0

39.7 38.3

10. CD player 66.6 66.7 66.5 11. None of the above 1.2 1.3 1.2 12. No response 0.7 1.1 ＞ 0.2

Question 31B: Which of the following media equipment do you have in your home? (Circle one or more options)


1. MD player 36.6 36.0 37.3 %

2. DVD player 65.3 65.7 64.9 3. HDD (Hard Disc Recorder) 37.0 38.8 35.1 4. Mobile telephone/PHS 85.8 86.5 85.1 5. PDA 15.5 16.4 14.5 6. Television capable of receiving CS digital

broadcasting 24.9

27.8 ＞ 22.1

7. Flat-screen television (LCD, plasma) 59.5 60.9 58.1 8. Digital camera 65.1 65.0 65.2 9. Video camera 33.3 33.5 33.0

10. Vehicle television (car navigation system, etc.) 40.2 39.7 40.6 11. Devices which enable 1seg viewing 30.9 30.4 31.5 12. None of the above 2.5 2.5 2.5 13. No response 2.5 2.8 2.1

39

◆ Sample distribution

Survey mode

Total Men/Women Age group Men age group Women age group

Men Women 16－39 40－5960 and older

16－39 40－5960 and older

16－39 40－5960 and older

Mail All

1,215 people

531 684 388 382 445 169 164 198 219 218 247

100.0% 43.7 56.3 31.9 31.4 36.6 13.9 13.5 16.3 18.0 17.9 20.3

Mail A

609 people

265 344 194 198 217 85 86 94 109 112 123

100.0% 43.5 56.5 31.9 32.5 35.6 14.0 14.1 15.4 17.9 18.4 20.2

Mail B

606 people

266 340 194 184 228 84 78 104 110 106 124

100.0% 43.9 56.1 32.0 30.4 37.6 13.9 12.9 17.2 18.2 17.5 20.5

Survey mode

Total

Occupation Agriculture/ forestry/ fishing

Self- employed

Sales/ service

Crafts/skilled labor

Office/technician

Executive/Manage-

ment

Specialist/freelance

/other

House- wife

Student UnemployedNo

response

Mail All

1,215 people

32 89 109 144 236 48 54 232 68 189 14

100.0% 2.6 7.3 9.0 11.9 19.4 4.0 4.4 19.1 5.6 15.6 1.2

Mail A

609 people

16 45 49 75 117 27 34 126 30 83 7

100.0% 2.6 7.4 8.0 12.3 19.2 4.4 5.6 20.7 4.9 13.6 1.1

Mail B

606 people

16 44 60 69 119 21 20 106 38 106 7

100.0% 2.6 7.3 9.9 11.4 19.6 3.5 3.3 17.5 6.3 17.5 1.2

Survey mode

Total

Size of city Region Special wards

& cities of 1 million or more

Cities of 300

thousand or more

Cities of 100

thousand or more

Munici-palities of

50 thousand or more

Munici-palities of less than

50 thousand

Hokkaido/Tohoku

KantoChubu/

Koshinetsu Kinki

Chugoku/Shikoku/Kyushu

Mail All

1,215 people

228 274 308 185 220 160 369 250 200 236

100.0% 18.8 22.6 25.3 15.2 18.1 13.2 30.4 20.6 16.5 19.4

Mail A

609 people

111 140 151 96 111 78 185 121 106 119

100.0% 18.2 23.0 24.8 15.8 18.2 12.8 30.4 19.9 17.4 19.5

Mail B

606 people

117 134 157 89 109 82 184 129 94 117

100.0% 19.3 22.1 25.9 14.7 18.0 13.5 30.4 21.3 15.5 19.3

40

◆ Response rate

Total Men/Women Age group Men age group Women age group

Men Women 16－39 40－59

60 and older

16－39 40－5960 and older

16－39 40－5960 and older

Mail All

Issued samples

1,800 people

867 933 651 550 599 313 275 279 338 275 320

Valid samples

1,215 people

531 684 388 382 445 169 164 198 219 218 247

Response rate

67.5% 61.2 73.3 59.6 69.5 74.3 54.0 59.6 71.0 64.8 79.3 77.2

Mail A

Issued samples

900 people

427 473 330 277 293 163 130 134 167 147 159

Valid samples

609 people

265 344 194 198 217 85 86 94 109 112 123

Response rate

67.7% 62.1 72.7 58.8 71.5 74.1 52.2 66.2 70.2 65.3 76.2 77.4

Mail B

Issued samples

900 people

440 460 321 273 306 150 145 145 171 128 161

Valid samples

606 people

266 340 194 184 228 84 78 104 110 106 124

Response rate

67.3% 60.5 73.9 60.4 67.4 74.5 56.0 53.8 71.7 64.3 82.8 77.0

Total

Size of city Region

Special wards

& cities of 1 million or more

Cities of 300

thousand or more

Cities of 100

thousandor more

Munici-palities

of 50 thousandor more

Munici-palities of less than

50 thousand

Hokkaido/ Tohoku

Kanto Chubu/

Koshinetsu Kinki

Chugoku/Shikoku/Kyushu

Mail All

Issued samples

1,800 people

372 408 432 276 312 216 588 336 288 372

Valid samples

1,215 people

228 274 308 185 220 160 369 250 200 236

Response rate

67.5% 61.3 67.2 71.3 67.0 70.5 74.1 62.8 74.4 69.4 63.4

Mail A

Issued samples

900 people

186 204 216 138 156 108 294 168 144 186

Valid samples

609 people

111 140 151 96 111 78 185 121 106 119

Response rate

67.7% 59.7 68.6 69.9 69.6 71.2 72.2 62.9 72.0 73.6 64.0

Mail B

Issued samples

900 people

186 204 216 138 156 108 294 168 144 186

Valid samples

606 people

117 134 157 89 109 82 184 129 94 117

Response rate

67.3% 62.9 65.7 72.7 64.5 69.9 75.9 62.6 76.8 65.3 62.9

41

◆ Reason for nonresponse

Total Total

nonresponsesNon-

returned

Change of address/address

unknown

Refused Illness Returned

blank Discrepancy in gender/age

Response by proxy

Other

Mail All

1,800 people

585 500 10 3 11 3 29 26 3

100.0% 32.5 27.8 0.6 0.2 0.6 0.2 1.6 1.4 0.2

Mail A

900 people

291 248 4 2 5 2 14 13 3

100.0% 32.3 27.6 0.4 0.2 0.6 0.2 1.6 1.4 0.3

Mail B

900 people

294 252 6 1 6 1 15 13 0

100.0% 32.7 28.0 0.7 0.1 0.7 0.1 1.7 1.4 0.0

characteristics of response tendency in mail surveys · gave the following summary regarding middle...

Documents