Table 4 Results of two sample one-sided... | Emerald Publishing

Table 4

Results of two sample one-sided t-tests with dependent variable being decline and recovery, i.e. turnover change 2019–2021

Nominal independent variables	Ownership (public, Private)	Size (larger, Smaller)	Line-of-business (freight, passenger and Freight)	Specialization (multipurpose, Single-purpose)	Freight direction (export, import, Bi-directional)	Freight type (bulk, unitized + Both)
t-test	None (there were only two private ports)	Ports divided into two equal-sized groups: larger (n = 9) and smaller (n = 9) based on their 2019 turnover	Testing the difference between two groups: Freight (n = 14), Passenger and Freight (n = 4)	Testing the difference between multipurpose (MP, n = 12) and single-purpose ports (SP, n = 6)	Testing the difference between export-oriented ports (n = 9) and bi-directional ports (n = 6); import-oriented not tested	Testing the difference between bulk ports (n = 9), and ports handling only unitized and both types of cargo (n = 9)
Hypothesis		H₁ = smaller ports recovered better than larger ports, i.e. their turnover change 2019–2021 was higher	H₁ = ports handling both passengers and freight recovered better than freight-only ports (their turnover 2019–2021 change was higher)	H₁ = SP ports recovered better than MP ports, i.e. their turnover change 2019–2021 was higher	H₁ = export-oriented ports recovered better than bi-directional ports, i.e. their turnover 2019–2021 change was higher	H₁ = Bulk only ports recovered better than ports with unitized and bulk cargo, i.e. their turnover 2019–2021 change was higher
Result		H₀ rejected (p = 0.014); smaller ports recovered better, with mean turnover change 2019–2021 of 3.3%, compared to −11.6% for larger ports	H₀ not rejected, p = 0.298; mean turnover change of freight ports was −5.08% and −1.089% for passengers and freight ports	H₀ not be rejected (p = 0.104); mean for MP ports was −7.54% and for SP ports 2.49%; however, this was borderline significant	H₀ not rejected (p = 0.300); mean turnover change was −6.88% and for bi-directional ports −10.31%	H₀ rejected at risk (p = 0.049); bulk-only ports recovered better; with mean turnover change of 1.9 compared to −10.3% for other ports
Diagnostics		t-test statistic = −2.411, p = 0.014, effect size is large (1.14); Shapiro–Wilk test assumed normal distributions; F test showed equal variance (p = 0.16)	F-test indicated equal variances (p = 0.286)	Although H₀ was not rejected, the effect size was large; F-test confirmed equal variance	F test indicated equal variance (p = 0.35)	t-test statistic = −1.762, p = 0.049; Shapiro–Wilk confirmed normality for both groups

Nominal independent variables	Ownership (public, Private)	Size (larger, Smaller)	Line-of-business (freight, passenger and Freight)	Specialization (multipurpose, Single-purpose)	Freight direction (export, import, Bi-directional)	Freight type (bulk, unitized + Both)
t-test	None (there were only two private ports)	Ports divided into two equal-sized groups: larger (n = 9) and smaller (n = 9) based on their 2019 turnover	Testing the difference between two groups: Freight (n = 14), Passenger and Freight (n = 4)	Testing the difference between multipurpose (MP, n = 12) and single-purpose ports (SP, n = 6)	Testing the difference between export-oriented ports (n = 9) and bi-directional ports (n = 6); import-oriented not tested	Testing the difference between bulk ports (n = 9), and ports handling only unitized and both types of cargo (n = 9)
Hypothesis		H₁ = smaller ports recovered better than larger ports, i.e. their turnover change 2019–2021 was higher	H₁ = ports handling both passengers and freight recovered better than freight-only ports (their turnover 2019–2021 change was higher)	H₁ = SP ports recovered better than MP ports, i.e. their turnover change 2019–2021 was higher	H₁ = export-oriented ports recovered better than bi-directional ports, i.e. their turnover 2019–2021 change was higher	H₁ = Bulk only ports recovered better than ports with unitized and bulk cargo, i.e. their turnover 2019–2021 change was higher
Result		H₀ rejected (p = 0.014); smaller ports recovered better, with mean turnover change 2019–2021 of 3.3%, compared to −11.6% for larger ports	H₀ not rejected, p = 0.298; mean turnover change of freight ports was −5.08% and −1.089% for passengers and freight ports	H₀ not be rejected (p = 0.104); mean for MP ports was −7.54% and for SP ports 2.49%; however, this was borderline significant	H₀ not rejected (p = 0.300); mean turnover change was −6.88% and for bi-directional ports −10.31%	H₀ rejected at risk (p = 0.049); bulk-only ports recovered better; with mean turnover change of 1.9 compared to −10.3% for other ports
Diagnostics		t-test statistic = −2.411, p = 0.014, effect size is large (1.14); Shapiro–Wilk test assumed normal distributions; F test showed equal variance (p = 0.16)	F-test indicated equal variances (p = 0.286)	Although H₀ was not rejected, the effect size was large; F-test confirmed equal variance	F test indicated equal variance (p = 0.35)	t-test statistic = −1.762, p = 0.049; Shapiro–Wilk confirmed normality for both groups

Source(s): Table created by authors