Sample: nature and size

Knowledge base

Sample: nature and size

The NTS response was getting lower. This is why the new design was started in 1999.

Response old NTS
Year	Accessible by telephone	CATI Response	Diary Response	Total¹
1985	83.0%	75.0%	81,5%	50,7%
1986	82.3%	76.0%	77,9%	48,7%
1987	82.3%	75.5%	77,6%	48,2%
1988	81.7%	75.0%	72,9%	44,7%
1989	81.4%	77.0%	75,6%	47,4%
1990	81.0%	75.4%	73,2%	44,7%
1991	80.9%	75.6%	75,6%	46,2%
1992	78.1%	76,4%	74,6%	44,5%
1993	77.3%	76.2%	73.1%	43.0%
1994	75.4%	72.6%	75.4%	41.3%
1995	75.1%	72.0%	74.6%	40.3%
1996	73.4%	71.1%	73.8%	38.5%
1997	73.8%	69.9%	71.3%	36.8%
1998	72.7%	68.0%	70.2%	34.7%
Response new NTS
Year	Accessible by telephone	CATI Response	Diary Response	Total
1999	80.6%	79.6%	36.1%	71.2%
2000	79.6%	80.0%	33.6%	70.4%
2001	78.1%	78.5%	33.4%	68.5%
2002	79.7%	77.0%	31.4%	67.7%
2003	77.7%	78.9%	31.4%	68.3%

Table. Response NTS. For old NTS the total column is the product of the 3 columns. For the new NTS, the percentages refer to different subgroups and the total response cannot be derived from the numbers shown.
¹The completeness of the percentage responding households is only 85%.

The table below gives the sample size.

	Households	Individuals	Journeys	Journey parts
1985	9287	25348	73709	79593
1986	10154	27253	77903	84355
1987	10170	26753	77080	83876
1988	10024	25782	75027	81212
1989	10270	26479	79047	85914
1990	10139	25552	78029	85092
1991	9569	23900	70513	77511
1992	10235	25546	75373	82580
1993	10029	24817	72773	80010
1994	34454	82835	298138	329493
1995	68433	167923	599453	657691
1996	62785	152547	542032	595628
1997	60124	143979	513656	564764
1998	58082	137991	486309	535660
1999	63336	146348	465358	500050
2000	64240	146528	455694	487635
2001	57491	130146	401070	429821
2002	44058	100266	309260	332151
2003	34454	77317	239078	254563

Table. Size of the NTS database.

Up to 1998

Taking province and degree of urbanisation into account, the sample was drawn from the Geographic Basic Register. The addresses to be phoned were randomly distributed over the days of the year.
In the 1985-1993 period, about 16,000 addresses were phoned. About 76% of the households were prepared to cooperate in the survey. Of the 26,000 diaries sent, 20,000 (76%) useable diaries were received. If we assume that the average household size of the response and non-response were about the same, the net response was about 58%.
In 1994, the national sample was twice as large; extra diaries were distributed in the transport regions of Utrecht and Arnhem/Nijmegen. In 1995, the national sample was enlarged further, to about 70,000 households per year.
This increase led to a better insight in the regional spread of journeys and the use of public transport.
Since 1994, children younger than 12 years old also took part in the survey. Moreover, the sloped was distinguished from the moped, and respondents were asked if they had a Public Transport Year Pass.
As a result of the method used, no households took part in the survey if it had/was:

no telephone,
an unlisted telephone number,
a home or barracks,
an office or holiday address.

Households that were difficult to reach by telephone are probably under-represented.

Since 1999

In the New NTS, the sample is drawn from the Borough Basic Administration. Now those who live at the drawn address are written to by name, which leads to a better response.

Households without a telephone or with an unlisted number are also recruited to take part.

In 2002, 64,006 households were approached. Eventually, more than 44,000 household questionnaires and 100,000 individual questionnaires were processed. In 2003, these numbers were lower because CBS could process fewer questionnaires for the same amount of money .

In 2004, the ministry's Transport Research Centre AVV, being the largest customer, took over the NTS from CBS. It is now carried out under the name of the Netherlands Mobility Study (NMS). The take over is not supposed to cause a trend break. The method was adopted completely and the sample was slightly smaller than that in the NTS 2003.

Next...