|
idea and conception: Ulf-Dietrich
Reips
programming: Stefan
Stieger
online
at:
http://www.sclog.eu/
Task
Before you start
Task
Scientific LogAnalyzer's task is to data
mine and analyze log files, for example those produced by
web servers. Scientific LogAnalyzer filters out the information
contained in the files and organizing this information for
easy import into standard software like SPSS, Excel, etc.
(via tab-delimited file).
Universally useful, with special features
created to meet the needs of behavioural scientists, especially
experimental psychologists and Internet scientists: While
Scientific LogAnalyzer is useful for the analysis of any type
of log file, it is the only tool of its kind with features
geared to behavioural scientists, such as handling of factorial
designs, response time analysis, and dropout analysis.
Speed: The option of calculating the tab-delimited
file only serves the purpose of speedy analysis – with
this option turned on a typical log file analysis will take
only seconds.
No need to install anything: Through
the upload of the log file and the download of the analyzed
output file, Scientific LogAnalyzer can be used from any location
in the universe, as long as it is connected to the Internet.
Before
you start
Important! If sensitive information is involved,
please save the log file for the upload in a secure directory
(https access).
The server log files must have the correct
line feeds (DOS format: \n).
Pre-filtering: Because server log files
can get very large, it is necessary to prepare the log files
before commencing with the upload by extracting the relevant
log file lines (e.g. all log file lines with a given experiment
number or other identifying string). For example, this can
be done with the command "copy lines containing..."
in the program BBEdit or a similar feature in WinEdit. Consider
further reducing the size of the log file by cutting lines
that are unnecessary for your analysis, for example those
containing image extensions (.gif, .jpg, ..)
Important! All input is case sensitive!
Important! The characters "<"
and ">" may not be used in the Search Arguments.
These are filtered out by Scientific LogAnalyzer in order
to subdue the interpretation of any HTML codes contained in
the log file.
Requirements:
- JavaScript must be enabled in the web-browser
- Popup windows must be enabled (important for the presentation
of help texts)
Important! Only file names with the following
extensions will be presented for standard analysis (we will
add other extensions upon request):
.html
.htm
.php
.php3
.asp
.acgi
Important! Before commencing with the analysis,
please make sure that the log file is sorted according to
time and date, starting with the oldest.
Important! Avoid breaking off the
process while doing the analysis (e.g. by use of the back-button),
as this may cause errors in the output file (because, on the
server, the process runs on. If a new process is started,
the output file will contain two processes, meaning: redundant
data).
|