# Starting the analysis

- Once imported, verify the data on the data_form data sheet. The filter columns(V and W) will show you whether a person’s data are valid and whether or not they can be used for regressions based on characteristics related to personal qualifications only (Regression pq) and standard regressions (valid data records are marked with «1», invalid data records with «0»).
- Orange cells indicate entry errors. Correct the errors and add any missing data. Please note:
- In the particular case of the «Gross wage», «Allowances» «13th wage» and «Special payments» columns, cells are marked in orange if you have left them empty instead of entering the figure 0. The system does not exclude these persons from the analysis as long as you have indicated at least one amount in either the «Gross wage», «Allowances» or «Special payments» column (the system automatically allocates the value 0 to empty cells).
- Similarly, for the two columns regarding working hours, «Level of employment» and «Hours paid», the cells are marked orange if you have left them empty instead of putting the number 0. The system does not exclude these people from the analysis as long as an amount has been indicated in one of the two columns (the system automatically allocates the value 0 to empty cells) (reminder: for each row, only one of the two columns has to be completed with, depending on the case, an indication of the percentage or a number of hours; the other column should contain a 0).
- Data that have been imported in text format instead of number format are also marked in orange; you must convert these data into number format in the import file.
- The columns T to Y cannot and must not be filled in; they are calculated by Logib.
- Click on «Start analysis». Depending on the amount of data involved the analysis will take some seconds or even minutes.
- Please wait until a message appears telling you that the analysis is now finished. Confirm by clicking on «OK».
- Once the analysis has been carried out, you can revise data in the data_form table and start a new analysis. However, by doing this, you will lose the existing analysis. You should, therefore, first save the Logib file under a new name.

# FAQ - Starting the analysis

#### My computer keeps freezing. What is the problem?

Data import and the analysis only take a short time. If your PC is frozen, check that you have installed the most recent Service Packs available. Furthermore, using Logib with Beta or Test version of Excel is not recommended.

#### What should I do if I receive a message similar to the following?: “There are a total of 150 observations but not all are valid and therefore suitable for statistical calculations. Should you wish to cancel the analysis in order to first revise your data, click on “cancel“. Do you wish to continue with the analysis?”

Only valid rows are included in the regression calculation. You will receive this message even if just one row of data cannot be included.

It is advised to cancel the analysis and to correct or complete the cells marked in orange on the data_form data sheet. If the whole Date of birth or Years of service column is orange, you probably indicated the wrong format when you imported your Excel file (by selecting the format «year of birth, e.g. 1982» for example, if the data in your Excel file are in DD.MM.YYYY format).

Columns V and W state whether a row (or observation) is valid (1) or not (0). An observation can only be used if it contains all the data required for the regression calculation. For the regression pq these are:

- Key
- Year of birth
- Sex
- Years of service
- Training
- Level of employment or paid hours
- At least one of the following columns: gross wage, allowances or special payments

For the standard regression, the system also needs:

- Level of qualifications required for the job
- Professional position

From a technical point of view, the regression can be calculated from the following number of observations:

- Regression pq: 7 observations
- Standard regression: 16 observations

However, in order to obtain a statistically significant result, a minimum of 50 observations is required.

#### How can I delete rows of incorrect data?

It is not possible to delete whole data rows from the data_form sheet. However, you can leave incomplete data rows in the data_form sheet as these rows will be automatically identified by the system during the analysis (valid rows are identified by a 1 in the columns V and W, invalid rows by a 0) and are not included in the regression analysis. If all rows are incorrect, however, you will have to start the operation again and reimport data into a new Logib file, this time using the correct parameters.

### Contact

Patric Aeberhard

Tel. 058 462 68 42

Mail Patric Aeberhard

Logib Helpline

Tel. 0800 55 99 00

Mail Logib Support

### Link

### Documents

Printable version Logib step-by-step

Last modification: 28.07.2015

Size: 856 KB | Type: PDF