Appendix D. Arkansas Statistical Summaries
The following tables highlight statistical results for the Arkansas survey data. In each table, the number of respondents to the question is shown by n (e.g., n=555).
Table 22 presents the statistical test results of whether drivers agreed that because of the info on the electronic message signs, they were better prepared to react to slow or stopped traffic. A single sample one-tailed t-test with 95 percent confidence showed that the drivers agreed that the electronic message signs better prepared them to react to slow or stopped traffic.
| Dataset (n = 555) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs combined – drivers were better prepared to react to slower traffic with the DMS | 1.72 ± 0.05 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -25.60 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
Table 23 presents the statistical test results for drivers' agreement with the statement that DMS makes them feel safer. A t-test showed that the drivers did agree that the DMS made them feel safer traveling through this construction zone.
| Dataset (n = 555) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs combined – DMS makes drivers feel safer in the construction zone | 2.80 ± 0.07 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -2.86 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
Table 24 presents the statistical test results for drivers' agreement that the Web site improved trip planning. A t-test with 95 percent confidence level showed that the 23 drivers who used the Web site agreed that the Web site was able to improve trip planning.
| Dataset (n = 23) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – The info on the Web site improves my ability to avoid delay on this section of I-30 | 2.00 ± 0.29 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -3.45 |
| Tabulated t-value (p=0.05) | -1.717 |
| Outcome | Reject Null hypothesis at p=0.05 |
Tables 25 and 26 present the statistical test results for drivers' agreement that the DMS messages contained enough detail and that the signs were located in the right places. Using a t-test with 95 percent confidence, drivers indicated that they agreed that the DMS contained enough detail and were located in the correct places.
| Dataset (n = 552) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs Combined – DMS messages are detailed enough to help me make decision about my route | 1.78 ± 0.05 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -24.40 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
| Dataset (n = 542) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs Combined – DMS are located in the right places to help me make decisions about my route | 2.10 ± 0.06 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -15.00 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
Table 27 presents the statistical test results for drivers' agreement that the HAR messages were detailed enough to help them make decisions. The average rating for this question was 1.93, indicating that the drivers agreed (1 is completely agree; 2 is somewhat agree) that the HAR messages were detailed enough to help them make decisions about their routes.
| Dataset (n = 121) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – HAR messages are detailed enough to help me make decisions about my route | 1.93 ± 0.11 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -9.73 |
| Tabulated t-value (p=0.05) | -1.658 |
| Outcome | Reject Null hypothesis at p=0.05 |
Tables 28 and 29 present the statistical test results for whether drivers agreed that the dynamic message signs were easy to understand and detailed enough to help them make decisions. At a 95 percent confidence level, results were significant when drivers were asked if the DMS are detailed enough and easy to understand.
| Dataset (n = 548) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – DMS messages are easy to understand (only drivers that also answered "The messages are detailed enough to help me make decisions about my route." | 1.26 ± 0.03 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -58.00 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
| Dataset (n = 548) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – DMS messages are detailed enough to help me make decisions about my route (only drivers that also answered "The messages are easy to understand.") | 1.78 ± 0.05 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -24.40 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
Tables 30 and 31 present the statistical test results for whether drivers felt that the HAR broadcast quality was too poor to hear, and whether the messages were detailed enough to help them make decisions. The findings on whether HAR broadcasts were too poor to hear was significant at a 95 percent confidence level. However, drivers agreed that the messages were detailed enough to allow them to make better decisions about their routes.
| Dataset (n = 124) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – Radio broadcast quality is too poor to hear properly (only drivers that also answered ("HAR messages are detailed enough to help me make decisions about my route.") | 2.57 ± 0.14 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -3.07 |
| Tabulated t-value (p=0.05) | -1.658 |
| Outcome | Reject Null hypothesis at p=0.05 |
| Dataset (n = 120) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – HAR messages are detailed enough to help me make decisions about my route (only drivers that also answered "Radio broadcast quality is too poor to hear properly.") | 1.93 ± 0.11 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -9.73 |
| Tabulated t-value (p=0.05) | -1.658 |
| Outcome | Reject Null hypothesis at p=0.05 |
Table 32 provides the statistical test results for whether drivers agree that the information on the Web site is accurate. The average rating of the 23 drivers who answered the question on whether the Web site is accurate was a 1.7. This value falls between completely agree (1) and somewhat agree (2). Using a t-test with 95 percent confidence, the drivers agreed that the Web site was giving accurate information.
| Dataset (n = 23) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – The info on the Web site is accurate | 1.70 ± 0.25 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -5.20 |
| Tabulated t-value (p=0.05) | -1.717 |
| Outcome | Reject Null hypothesis at p=0.05 |
Table 33 presents the statistical results for whether drivers found the DMS messages to be accurate. Performing a t-test for the responses shows that the drivers found the DMS messages accurate at a 95 percent significance level. These findings support the conclusion that participants feel the DMS were giving accurate messages.
| Dataset (n = 543) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – DMS messages are accurate | 1.75 ± 0.05 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -25.00 |
| Tabulated t-value (p=0.05) | -1.645 |
| Outcome | Reject Null hypothesis at p=0.05 |
Tables 34, 35, and 36 provide the statistical test results for whether drivers found the HAR messages to be useful, accurate, and updated frequently enough. Using a t-test at 95 percent significance showed that the drivers agreed that HAR was useful, accurate, and updated frequently enough. The scale included completely agree (1), somewhat agree (2), neutral (3), somewhat disagree (4), completely disagree (5), and not sure.
| Dataset (n = 115) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – HAR messages are useful (only drivers that also answered "HAR messages are accurate," and "HAR messages are updated frequently enough.") | 1.74 ± 0.10 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -12.6 |
| Tabulated t-value (p=0.05) | -1.658 |
| Outcome | Reject Null hypothesis at p=0.05 |
| Dataset (n = 114) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – HAR messages are accurate (only drivers that also answered "HAR messages are useful," and "HAR messages are updated frequently enough.") | 2.00 ± 0.11 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -9.09 |
| Tabulated t-value (p=0.05) | -1.658 |
| Outcome | Reject Null hypothesis at p=0.05 |
| Dataset (n = 98) | Level of agreement (Mean ± Standard Error) |
|---|---|
| Private drivers & CVDs – HAR messages are updated frequently enough (only drivers that also answered "HAR messages are useful," and "HAR messages are accurate.") | 2.27 ± 0.14 |
| Statistical Comparison Results (Null Hypothesis: Mean of the dataset is equal to or greater than 3; Alternative Hypothesis: Mean of the dataset is less than 3) |
|
| Calculated t-value | -5.21 |
| Tabulated t-value (p=0.05) | -1.658 |
| Outcome | Reject Null hypothesis at p=0.05 |
Note: Of the 616 surveys, only two PVDs and three CVDs did not provide gender information, 7 PVDs and 13 CVDs did not give age information, and four CVDs did not give their driving experience. Table 37 presents the statistical test results for the question of whether the combined volume from the ITS detectors was significantly higher than the AHTD Tube data. A t-test showed that the combined volume from the ITS detectors was significantly different – higher - with a 95 percent confidence level.
| Dataset (n = 240) | Hourly volume difference between tube data and ITS detectors (in percentage) (Mean ± Standard Error) |
|---|---|
| May 21 – Jun 20, 2004. Detectors No 62 & 26 (RTMS), and data from I-30 & Raymar Rd (Tube data) | -42.4 ± 2.1 |
| Statistical Comparison Results (Null Hypothesis: Mean of the differences equals 0; Alternative Hypothesis: Mean of the differences does not equal 0) |
|
| Calculated t-value | -20.19 |
| Tabulated t-value (p=0.05) | 1.960 |
| Outcome | Reject the null hypothesis at p=0.05 |
Table 38 presents the statistical results for the question of whether the combined volume from the ITS detectors was different than AHTD's tube data. A t-test showed that the combined volume from the ITS detectors was not different than the AHTD data with a 95 percent confidence level. Tables 37 and 38 test significance for two different sources of ITS detector data.
| Dataset (n = 240) | Hourly volume difference between tube data and ITS detectors (in percentage) (Mean ± Standard Error) |
|---|---|
| May 21 – Jun 20, 2004. Detectors No 63 & 25 (RTMS), and data from I-30 & Raymar Rd (Tube data) | 0.071 ± 0.17 |
| Statistical Comparison Results (Null Hypothesis: Mean of the differences equals 0; Alternative Hypothesis: Mean of the differences does not equal 0) |
|
| Calculated t-value | 0.42 |
| Tabulated t-value (p=0.05) | 1.960 |
| Outcome | Fail to Reject null hypothesis at p=0.05 |
