You should check or uncheck the 'parse PDF' option based on the input file that you have selected.
If your file is a searchable PDF document and you check this option, then StatementReader will extract the text from the PDF document directly without evoking its internal OCR engine. This method often offers the best levels of speed and accuracy.
If your file is a searchable PDF document but you uncheck this option, then StatementReader will use its OCR engine to process your file. In very rare cases this may yield better results, but as a general rule (particularly where the PDF was created electronically without being scanned) it can be avoided.
If your file is not a searchable PDF document and you check this option by mistake, then StatementReader will report that no transactions have been found for each page.