China
Africa
Explore chapters and articles related to this topic
Insight
Published in Wanda Grimsgaard, Design and Strategy, 2023
The size of the population does not affect the size of the sample. You get the same margins of error whether you interview 1,000 in Norway with 3.5 million people or 1,000 in the US with 350 million people or China with 1.4 billion people. When you have a political opinion poll in Norway, it is often somewhere between 1.5%and 3% margin of error plus or minus. So, a party that gets e.g. 30% support will have a margin of error of around 3% points. If you say that a party has 30% support, then you assume that there is a 95% probability that the party is somewhere between 27% and 33% support. That is the margin of error. But the size of the margin of error is not affected by the size of the population. If you interview 1,000 voters in Norway (with about 3.5 million voters), you get a margin of error of 3%. If you interview 1,000 voters in the United States (with over 200 million voters), you also get a margin of error of 3% points, because the margins of error or the confidence of a sample survey are not affected by the size of the population (Thorsen, 2017).
Statistics
Published in Dušan Teodorović, Miloš Nikolić, Quantitative Methods in Transportation, 2020
Dušan Teodorović, Miloš Nikolić
We can also generate a few different samples (Figure 5.16). Let every sample contain n elements: x1,x2,…,xn. We could estimate the average parking duration that is based on the random sample of n vehicles. Since various samples generate different average parking durations, we conclude that there is variability in our estimate. We introduce into the analysis the margin of errors (the amount of “plus or minus”). This margin of errors measures the variability of the parameter estimate. The greater the sample size, the smaller the margin of error.
Data collection, processing, and database management
Published in Zongzhi Li, Transportation Asset Management, 2018
For the simple random sampling method, the following inputs help determine the proper sample size: The accuracy and precision level, measured by the maximum expected difference between the true population parameter and the corresponding sample estimate expressed by a probability statement (e.g., confidence level), also known as the margin of error (ME). For example, 85.5% of the local highway segments have PSI over 2. The ME is +5% with a confidence level of 90%. That means the probability of the event that the true percentage falls into 85.5% + 5% is 90%Significance level α. For an estimation problem, α is 1 − confidence levelCritical standard z-score. For an estimation problem, the critical standard score is the value for which the cumulative probability is 1 − α/2The population size NThe population variance σ2
What Motivates People to Adopt Personal Health Record? Understanding the Effect of Innovation and Context-Specific Threat on Personal Health Record Adoption Intention
Published in International Journal of Human–Computer Interaction, 2023
To find out the adequate sample size for evaluating the population prevalence with good precision, the following formula proposed by Nduneseokwu et al. (2017) is used. where S is the sample size, Zscore associates with the confidence level (95% confidence level was selected), p is the standard deviation (0.5 is set to ensure a large sample). The margin of error is associated with the confidence interval (±5%). In step one, as Nduneseokwu et al. (2017) suggested, we determine the sample size for infinite populations, given that Zscore is 1.96 for a 95% confidence level, resulting in a sample size of 385. In step two, we adjust the sample size of Taiwan to the definite population (23,900,579) to obtain the adjusted population size by using the formula (Nduneseokwu et al., 2017):
How accurate is your travel time reliability?—Measuring accuracy using bootstrapping and lognormal mixture models
Published in Journal of Intelligent Transportation Systems, 2018
Other measures of accuracy can be derived based on standard errors. For example, the margin of error is the standard error associated with a particular confidence level; double the width of the margin of error centered on the desired statistical estimator can be considered as a confidence interval. The Student's T-test is one of the applications of confidence intervals. Several previous studies (for example, Cambridge Systematics, Inc., 2012; Srinivasan & Jovanis, 1996; Toppen & Wunderlich, 2003) investigated the issue of sample size requirement for travel time measurement based on these measures of accuracy. However, standard errors (Eqs. 1 and 2) have two major disadvantages: Equations 1 and 2 assume a normal distribution, and accordingly the data should statistically follow the normal distribution.Explicit equations for measuring standard errors (e.g., Eqs. 1 and 2) do not exist for most statistical estimators (e.g., median, nth percentile, and standard deviation) other than the mean (Efron & Tibshirani, 1993).