# Exploring Statistical Analysis in Undergraduate Economics: A Sample Study

## Data Analysis

The key aim of this study was to determine the differences in the trends of long-term unemployment between Europe and the US. For that purpose, the panel data of 39 years of both the US and European countries have been gathered. This section presents the study’s results by analyzing the collected data through the GLS model. The results of descriptive statistics and correlation analysis are also presented in this section of the study.

## Descriptive Statistics

In Table 1, the results of descriptive statistics are highlighted. According to Mishra et al. (2019), descriptive statistics is considered one of the most useful techniques for quantitatively summarising or describing the key components of collected data. In the Table, some key features of collected data are outlined, like the number of observations, mean, and standard deviation for each variable included in the model.

Table 1 Descriptive Statistics

 Variable Obs Mean Std. Dev. Min Max Tradeunion~e 414 36.21329 23.3775 8.5 97.2 UnionCover~e 248 65.8996 29.89763 11.5 100 Employment~n 252 2.827036 .6433816 11.5 4.125 LTUNEMP 448 36.25093 15.85969 1.515152 73.49994

As table 1 of descriptive statistics mentions, the number of observations for each variable is different. The total number of observations for Trade Union density was 414, whereas for Union Coverage, the number of observations was 248. On the other hand, the number of observations for employment protection is recorded at 252, and for long-term unemployment, the number of observations is 448. The mean value presented in table 1 depicts that the average percentage of trade union density in European countries and the US is 36.213, which is deviated by 23.37.

Apart from that, the mean value of Union Coverage is computed at 65.89, which is deviated by 29.89. Employment protection has a mean value of 2.82 with a standard deviation of 0.643. This implies that the average employment protection in the US and European countries is 2.82, which can deviate by 0.643. Lastly, the average long-term unemployment in all the investigated countries is recorded at 36.25, which can deviate by 15.859.

## Correlation Analysis

The correlation analysis test has been performed to analyze the strength of the association between the variables. According to Schober, Boer and Schwarte (2018), the correlation analysis technique is widely used to assess the level of association between the variables. Anything below 0.3 is considered a weak association, whereas 0.3 to 0.7 is considered a moderate association.

On the other hand, a correlation value of more than 0.7 shows a strong association between the variables. As per the results of the correlation analysis presented in Table 2, all the variables, including employment protection, union coverage, and trade union density, have a moderate association with long-term unemployment.

Table 2 Correlation Analysis

 Tradeu~e UnionC~e Employ~n LTUNEMP Tradeunion~e 1.0000 UnionCover~e 0.4050* 1.0000 Employment~n -0.2724* 0.4993* 1.0000 LTUNEMP -0.3456* 0.3989* 0.5120* 1.0000

## GLS Model

The results of the GLS model are presented in table 3. The generalized least square (GLS) model is a useful method for assessing the unidentified parameters in a linear regression model when there is a certain extent of association among the residuals in a regression model (Arruda and Bentler, 2017). As per the results of the GLS model highlighted in table 3, all the independent variables, including trade union density percentage, employee protection, and union coverage, significantly impact long-term unemployment. On the other hand, the dummy variable of Europe also has a significant impact on long-term unemployment. This implies a significant difference between the European countries and the US in terms of long-term unemployment.

 Estimated covariances 1 Number of obs 145 Estimated autocorrelations 0 Number of groups 12 Estimated coefficients 5 Obs per group: Min 4 Avg 12.08333 Max 20 Wald chi2(9) 142.90 Prob > chi2 0.0000 Log Likelihood = -547.4994 LTUNEMP Coef. Std. Err. Z P>|z| [95% Conf. interval] TradeuniondensityPercentage -.2192386 .0542268 -4.04 0.000 -.3255212 -.1129561 UnionCoverage -.179443 .0599981 -2.99 0.003 -.2970371 -.0618488 EmploymentProtection 7.393195 1.956202 3.78 0.000 3.559109 11.22728 Europe 22.7477 3.842003 5.92 0.000 15.21751 30.27788 _cons 13.36089 5.011579 2.67 0.008 3.538374 23.1834

## References

Arruda, E.H. and Bentler, P.M., 2017. A regularized GLS for structural equation modelling. Structural Equation Modeling: A Multidisciplinary Journal24(5), pp.657-665.

Mishra, P., Pandey, C.M., Singh, U., Gupta, A., Sahu, C. and Keshri, A., 2019. Descriptive statistics and normality tests for statistical data. Annals of cardiac anaesthesia22(1), p.67.

Schober, P., Boer, C. and Schwarte, L.A., 2018. Correlation coefficients: appropriate use and interpretation. Anesthesia & Analgesia126(5), pp.1763-1768.