Content

This course covers some basic principles for designing experiments, topics related to linear models for an analysis of variance and analysis of covariance, and conducting an appropriate analysis of data from several types of experiments: completely randomized, randomized complete block, split plot and cross-over.

Modes of study

Participation in classroom work, exam.

Recommended preceding studies

MTTTP1 Introduction to Statistics or other basic course in Statistics.

Course Contents

Matrix basic operations, random number generation, cross-validation, Jackknife, Bootstrap, use of the R program

Modes of Study and Registration

Independent work, only for CBDA-students.

Please see the Moodle page for details and instructions.

Contact teaching (lectures and excercises) is only in Finnish. For details, see the teaching schedule of the finnish version (MTTTA14 Tilastotieteen matriisilaskenta ja laskennalliset menetelmät). The lectures are based on study material which is available in English.

Modes of study

- Lectures
- Exercises (independent work)
- Exam

Enroll on the course TIETA6 Tietorakenteet in NettiOpsu

Self studying, weekly excercises, practical work, and exam.

Knowledge about statistical methods and data analysis is of great importance in almost any field of research. In this course, general concepts of statistics will be provided so that the students can be able to independently carry out a small scale empirical research with the statistical software R.

Contents

1. Basic concepts (1 lecture)
   1. What is statistics?
   2. Statistical research and its phases of work
   3. Population and sample
   4. Types of variables and scales of measurement
2. Descriptive statistics (2 lectures)
   1. What is data?
   2. summary statistics (mean, median, quartiles, standard deviation and variance)
   3. Plotting the data (scatter plot, histogram, QQ-plot and boxplot)
   4. Contingency tables
3. Basics of statistical inference (4-5 lectures)
   1. Basic concepts of probability
   2. Probability distributions (Binomial and Normal)
   3. Basic concepts in sampling theory
   4. Confidence intervals (mean and difference of means in independent samples)
   5. Hypothesis testing (mean and equality of means in independent samples, type I and type II-errors and p-values)
   6. Introduction to some non-parametric tests
4. Linear models (3 lectures)
   1. Introduction to linear regression
   2. Basics of logistic regression
   3. Basics of Analysis of Variance (ANOVA)

A maximum number of 50 students will be allowed in this course (70% doctoral students and 30% masters students).

Please note that this course cannot be included inside the minimum 120 ECTS of Master's Degree Programme in CBDA (basic level course).

MTTTP1 Tilastotieteen johdantokurssi lectured in period I, II or III-IV is recommended for Finnish students.

Content

• Lecture I (75 minutes): Estimation based on (i) Complete Sample, (ii) Failure-Truncated Sample; Quiz I (30 minutes): Based on Lecture I.
• Lecture II (60 minutes): Estimation based on Time-truncated Sample; Quiz II (30 minutes): Based on Lecture II.
• Lecture III (60 minutes): Prediction of Future Observation(s) Based on Exponential Samples; Quiz III (30minutes): Based on Lecture III.

Modes of study

Participation in course work and a practical work.

Please also note visiting lecture:

Professor (retired), Bikas Sinha, Indian Statistical Institute

Topic: “F AM NOT LICKED” – The Twelve Penny Problem with Applications

• 2.11.2017 at 14.15–15.00, Pinni B3109.

Are most published research findings false (Ionnadis, 2005, Open Science Collaboration, 2015)? The course deals with the interpretations of and possible solutions to the lack of replicating results in empirical research (e.g. social psychology and cancer research). As a result of the lack of successful registered replications, a growing number journals such as Psychological Science are favouring practices such as direct replications and registered reports. The course gives a hands-on introduction to two methods which are thought to improve the reliability and replicability of empirical research: p-curve analysis and pre-registration.

The course:

• Introduces participants to the ongoing conversation about the replicability of empirical research
• Introduces participants to different interpretative frameworks of p-values and statistical significance (Null Hypothesis Significance Testing, Fisher, Neyman-Pearson, Bayes)
• Provides participants with skills for pre-registration and p-curve analysis

Masters Degree students of CBDA-programme: course can be included in the advanced studies of CBDA (Statistical Data Analytics). For details, contact your Personal Study Plan teacher.

This course is organized as a web course with an introductory lecture.

Visualization of quantitative data when reporting and publishing findings

Course description:

It is commonly said that “a picture is worth a thousand words”. The same is true when reporting findings of an analysis of quantitative data. A proper visualization of the results might make the difference between the success and failure in telling a story or in publishing one’s findings. This course gives, first, a brief introduction to the R software and to D3 JavaScript library for manipulating documents based on data. Second, the course focuses on visualization of quantitative data which is of utmost importance when reporting and publishing findings. Examples and applications will be done for multivariate, temporal, spatial and text data. Examples used during the course will be based on the R software, and preliminary knowledge of this software is advisable but nor required. Participants can learn more about the R software prior to the course online at: http://www.uta.fi/cast/events/Ronline.html.

Goals: The course:

• Introduces participants to the R environment and to D3
• Introduces the participants to data exploration and data visualization
• Provides a recent variety of techniques and strategies to visualize quantitative data

Place: Computer classroom Ml 50 Linna

Programme

Day 1: 3.11.2017

09.15-12.00   Introduction to R

12.00-13.00   Lunch break

13.00-16.00   Introduction to data exploration and data visualization: types of data and of databases; online databases; Visualization of multivariate data

Day 2: 10.11.2017

09.15-12.00   Visualization of temporal data and of spatial data; text visualization

12.00-13.00   lunch break

13.00-16.00 Introduction to D3; Real time big data applications

PLEASE NOTE: Attendance to BOTH days is required for the completion of the course.

Teacher: Paulo Canas Rodrigues

Pre-assignment: Please write a short (one A4) text stating:

1)    Your name & disciplinary background

2)    State your own motivation for participating on this course and what do you expect to learn.

DEADLINE for the pre-assignments to be announced.

In addition, participants will write a mini-assignment after the second meeting with a two weeks’ deadline.

Enrolment via NettiOpsu. Maximum number of students is 24. Selection method is draw. Students should check the selection result via NettiOpsu after the enrolment period.

Modes of study

- Lectures
- Exercises (independent work)
- Exam

This course is organized as a web course.

R is one of the most widely used software for statistics and data science. In this course, the students will have the chance to become familiar with basic R objects, operations, visualizations tools and programming. The course will have in-class and online components. The online material (more information here: http://www.uta.fi/cast/events/Ronline.html) should be done before the first day of the course or between the two in-class days of the course. The second in-class day of the course will offer the chance to get a deeper knowledge about R and to analyze own data.

Modes of study: pre-assignment, participation in course work

BOTH online and in-class parts are required for the completion of the course.

 Pre-assignment: Please write a short (half A4) text stating:

1)    Your name & disciplinary background

2)    State your own motivation for participating on this course and what do you expect to learn.

Maximum number of students is 20. Priority will be granted for the first enrolments, based on the proportions: 60% PhD students 30% MSc students and 10% BSc students. Students should check the selection result via NettiOpsu after the enrolment period.

Multilevel models are designed to explore and analyse data that come from populations which have a complex structure. Behavioural and social data commonly have a nested structure. Multilevel models are becoming an increasingly popular method of analysis for situations where responses are grouped, such as in schools or other institutions, neighbourhoods, firms, parliamentary constituencies, or any other social or spatial clusters. A particular version of multilevel modelling is where there are multiple measures on each respondent, so that the grouping is of measures within person; where these multiple measures are taken on successive occasions, multilevel modelling provides a means of modelling individual change over time. This course will emphasise the practical application of multilevel models and lectures will be combined with practical sessions in order to reinforce concepts.

Course contents

• Duration (2 days)
• Lectures (6 h)
• Hands-on exercises with MLM tools (6 h)

Flash presentation on application of MLM in participant’s own research or field (5-10 minutes each)

Target group
The course is intended for post-graduate candidates and researchers who are interested in the use of multi-level modelling in their research. Understanding of basic statistics is required.

Enrolment: At the maximum 24 students. Priority will be given to those who need MLM in their research (PhD researchers most especially).

For pre-selection evaluation, please write a short text stating:

• Your name and disciplinary background
• Your motivation for participating in the course and what you expect to learn

Content

It is quite common that we do not get all the information we want for our statistical analysis. For example, in medical research a person can refuse to provide certain information they feel sensitive, such as weight, substance abuse, sexual orientation etc. Particularly, missing data in longitudinal studies is more the rule than an exception. Missing data in statistical analysis causes all sorts of problems. For example, the desired statistical method cannot be directly applied; loss of information or the results obtained can be biased if the analysis is not done properly accomplished. The course introduces various missing data mechanisms and their effects on statistical analysis. In addition, it presents and evaluates some of the commonly used methods for statistical analysis with missing data. Also special methods for the analysis of longitudinal data are presented including likelihood-based methods and multiple imputation.

Modes of Study

Course work, exam.

Ilmoittautuminen NettiOpsussa.

The course is organized as a web course with an introductory lecture.

Learning outcomes
After completing the course, the participants
- know the phases of the process of knowledge discovery (data prepocessing, data mining and postprocessing)
- know basic data mining tasks and methods
- are aware of possibilities of utilising data mining in different research fields

Description

In data mining, large quantities of data are explored and analysed by automatic and semi-automatic means to discover novel, interesting information. Data mining is an interdisciplinary field combining e.g. methods from computer sciences and statistics. It has wide, diverse application areas from education, social, business and administrative sciences to medical and life sciences.

Course contents
-    Lectures 10 h
-    Hands-on exercises with data mining tools 10 h
-    Reading research articles related to applications of data mining methods in participant’s own field and writing a short report
-    Giving a presentation on applications of data mining in participant’s own field (presentation session 3 h)

Teachers: Kati Iltanen, Martti Juhola, Henry Joutsijoki

Target group
The course is intended for post-graduate students who are interested in data mining. No computer sciences or statistics background is required.

Enrolment: At the maximum 15 students. Selection method is draw.

Teaching:

Lectures:
Wed 2.5.2018 at 10-12 Pinni B1083
Fri 4.5.2018 at 10-12 Pinni B1083
Wed 9.5.2018 at 10-12 Pinni B1083
Fri 11.5.2018 at 10-12 Pinni B1083
Wed 16.5.2018 at 10-12 Pinni B1083

Practices:

Wed 2.5.2018 at 12-14 computer classroom Pinni B1084
Fri 4.5.2018 at 12-14 computer classroom Pinni B1084
Wed 9.5.2018 at 12-14 computer classroom Pinni B1084
Fri 11.5.2018 at 12-14 computer classroom Pinni B1084
Wed 16.5.2018 at 12-14 computer classroom Pinni B1084

Presentation session
Fri 25.5.2018 at 10-13 Pinni B1083


Evaluation: Pass/fail

Learning outcomes
After the course, the student:
•    has developed an understanding of the nature of bibliometric research
•    is familiar with the most commonly utilised sources of bibliometric data and ways of applying this data in (daily) research conduct
•    is able to collect bibliometric data and to apply this data in various research settings
•    is able to utilise bibliometric tools for the benefit of his/her thesis writing and future career building
•    understands the limitations of bibliometric data and methods

General description
First, bibliometrics is a methodological approach in which the scientific literature itself becomes the subject of analysis. Bibliometrics offers a powerful set of tools that help scholars to, for example:

• delineate samples for literature reviews

• understand scientific citations

• investigate the research impact of authors, journals and institutes

• identify the most influential works of a given topic

• map the cognitive and organisational characteristics of research fields

• describe scientific networks within the knowledge base of distinct disciplines

• produce indicators for use in policy and management contexts

• understand the basics of (statistical) meta-analysis

As such, bibliometrics can be considered as the science of science.
Second, making an impact in the scientific community is a must for young researchers aiming for building a career at the university. Bibliometrics provides researchers a means of promoting and monitoring their own research impact.
This course focuses on the general features of bibliometrics 1) as a pivotal tool for conducting research and 2) as an integral part of building a research career.

Teaching schedule: 5.3. and 6.3.2018 at 10–16 o'clock

Teacher: Dr. Teemu Makkonen

Place: Computer Classroom Ml 50 (Linna building)

Completion: Those accepted to the course are required to send in a pre-assignment through Moodle. During the lectures and exercises, active participation is required. After the course, every student will write a short essay on how to apply bibliometrics in connection to their own scientific field and research topic.

Evaluation: Pass/Fail.

Enrolment: In NettiOpsu. Number of participants 20 at the maximum. Selection method is draw. Students should check the selection result from NettiOpsu after the enrolment period.

Course pre-assignment:
Before the course, course participants are expected to write a short summary-description (max 2–3 pages) of their own doctoral research by introducing:
1)    the topic of their research,
2)    the discipline and specific research field they are engaged in
3)    the “keywords” of their thesis, significant works (books, book chapters, articles) they are referring to in their research
4)    (if applicable) a list of their own published work.
Detailed instructions will be made available to enrolled students through Moodle.