Practicals

1. Class Admin, evaluation, midterms

EVERY STUDENT MUST BE SIGNED UP FOR THE LECTURE AND ONE PRACTICAL CLASS AS WELL!!!!

2. How to reach SPSS from home? 

https://iig.elte.hu/en/content/about-spss-program.t.17309

Each device must be linked into the university network in order to use the program. VPN connection is required for activation and for later use of program from home

3. Prepare a research plan - ppt slides

4. A small experiment: https://www.psytoolkit.org/experiment-library/go-no-go.html 

5. Opening SPSS, what do we see, the windows of SPSS

6. Creating databases in SPSS

   1. Entering data by hand.

   2. Opening already created SPSS databases

   3. Opening Excel databases in SPSS

7. Overview of Class1

Files for class1: StatisticsPracticalTopic02_MidtermGrades

1. Practice last class's material: 

   1. Open the StatisticsPracticalTopic2_MidtermGrades file in spss

   2. Give labels to the variables.

2. Descriptive statistics for the Sex variable

   1. mean, standard deviation, median, minimum, maximum

3. Frequency statistics for the Sex variable

4. Do a descriptive and frequency analysis for the Grades variable.
   What is the mean and standard deviation on this variable? 

5. Close this dataset

6. Task2: I collected data on 10 students anxiety score. The first 5 students were males, the second 5 students were females. The anxiety scores were as follows: 4, 10, 2, 7, 8, 3, 5, 5, 9, 3. Create this database in SPSS

7. Calculate the frequency for the gender data.

8. Calculate the mean anxiety score of my sample. 

1. Practice last class's material: 

   1. Open the StatisticsPracticalTopic2_MidtermGrades file in spss

   2. What is the mean and standard deviation for the grades variable? 

2. Task2: I collected data on 10 students reaction time data on a Stroop task. The first 5 students were females, the second 5 students were males. The RT were as follows: 30, 32, 40, 39, 41, 35, 32, 30, 39, 40miliseconds. Create this database in SPSS

   1. What is the mean and standard deviation for the RT variable? 

   2. normal curve, distribution of data

   3. skewness, kurtosis:
      Skewness is a measure of symmetry, or more precisely, the lack of symmetry. A distribution, or data set, is symmetric if it looks the same to the left and right of the center point. Kurtosis is a measure of whether the data are heavy-tailed or light-tailed relative to a normal distribution.

As a general rule of thumb: If skewness is less than -1 or greater than 1, the distribution is highly skewed. If skewness is between -1 and -0.5 or between 0.5 and 1, the distribution is moderately skewed. If skewness is between -0.5 and 0.5, the distribution is approximately symmetric. 

Kurtosis value can reach from 1 to + infinity. But generally, a kurtosis value = 3 (Mesokurtic) indicates a normal distribution. Kurtosis value > 3 indicates positive kurtosis (Laptokurtic - has tails that decay more slowly than a Gaussian, resulting in more outliers ) with low SD and Kurtosis value <3. indicates negative kurtosis (Platicurtic - which doesn’t necessarily have a flat top but produces fewer or less extreme outliers than the normal distribution. ). 

3. I also collected from the same set of participants RT data on a word naming task. Their RT were as follows: 100, 102, 98, 99, 100, 105, 103, 104, 110, 101 second.

4. Standardization of variables

1. open the StatisticsPracticalTopic02_ReactionTimeData datafile. 

2. Check the descriptive statistics of the variables. 

3. One sample t-test:
   I have read in a research paper done by XY, that the mean RT on the stroop task is 160msec. My hypothesis: I have a similar mean score on my current database on the stroopRT variable. 

To test the hypothesis that the mean score on my sample will be similar to previous studies, I ran a one sample t-test, and found t(15) = -3.005, p = 0.009. Based on the results, we could not confirm the hypothesis as our sample scored significantly higher than the compared value (mean=135.56, SD=32.52).

4. One sample t-test:
   I have read in a research paper that the meanRT on average in word naming an American sample is 200msec. My hypothesis: I have a similar mean score on my current database. 

To test my hypothesis, I ran a one sample t-test (t(15)=0.47, p=0.63). The mean in from our sample was 200.87ms, SD=7.18. The results suggest the means are not significantly different, hence, we can confirm our hypothesis. 

5. Practice tasks for the one sample t-test: 

   1. Literature results suggest a mean stroopRT reaction time of 130ms for males. Is it true on the current sample? 

   2. Literature results suggest a mean stroopRT reaction time of 150ms for females. Is it true on the current sample? 

   3. Literature results suggest a mean word naming RT reaction time of 190ms for females. Is it true on the current sample? 

   4. Literature results suggest a mean word naming RT reaction time of 175 ms for males. Is it true on the current sample? 

1. I have collected data on participants' novelty seeking with a questionnaire. I have collected data from 10 males, and 10 females. The scores for the males were as follows: 20, 22, 21, 20, 21, 23, 24, 20, 21, 21. The scores for the females were as follows: 25, 26, 23, 25, 25, 24, 26, 27, 26, 27.

2. Calculate descriptive statistics for these scales. 

3. Is there a significant difference in the mean novelty seeking score between males and females? H0: the means are similar. Independent sample t-test
   Solution example: I have run an independent sample t test to test the hypothesis. The Levene's test for equality was not significant, suggesting equal variances. The t-test results showed a significant difference: t (18) = -7.043, p <0.001. Mean score of the males were lower (21.2, Std 1.3) as compared to females (25.4, std 1.3). Males showed significantly lower novelty seeking as compared to females in the present sample. 

4. 5 years later, I repeated the same data collection, with the same set of participants. Their scores now were as follows: males: 26, 27, 23, 22, 25, 26, 26, 27, 28, 30. For the females: 24, 25, 26, 26, 26, 25, 27, 27, 26, 26. 

5. Was there a significant change in the mean novelty seeking scores of males or females? H0: the means are similar. Paired sample t-test. 

Run the analysis for females
Solution example: I run a paired sample t-test to test this hypothesis. The results showed no significant different: t (9) = -0.100, p = 0.343. For the 10 female participant, the mean score of novelty seeking was 25.4, std 1.26 at data collection1, and 25.8 std 0.91 at data collection2. These results suggest that there are no significant differences in the average novelty seeking scores of females in these 5 years. 

5. 1. Run the analysis for males. 

6. Practice: 

   1. Is there a significant difference in the mean novelty seeking scores of males and females at the second time of the data collection?

   2. My colleagues have also measured novelty seeking on a large representative sample of Hungarian participants, and they published that the mean score of novelty seeking on Hungarian participants is 22. Is there a significant difference between their published mean scores and the mean scores I measured at data collection 1 or 2?

• Cearte a database in spss based on the tables below. 

• In this database: 

  1. Is the mean anxiety score in the total present sample significantly higher as compared to the depression score? 

  2. In the total sample, do females score higher on the anxiety score than males? 

  3. We found in our previous study that the mean score on a large, Hungarian students is 9.80 for the anxiety scale and 3.75 for the depression scale. Do the present sample show significant differences than the earlier study in the mean scores of the scales? 

  4. Do females show significantly higher scores on the depression or on the anxiety scores? - confirm your results with statistical analysis. 

  5. Let's calculate out the mean 'mood characteristic' scores on the sample, by taking the average of the anxiety and depression scores. Who scores higher in this new mood characteristics score - females or males? Is the difference significant?

  6. Do males score higher on the depression score than females? 

  7. Do St. Catherine or St. Celemnts school students score significantly higher on the mood characteristics score? 

  8. My hypothesis is that student at St. Margit school have higher anxiety scores than students of the St. Celement school. Is it true on the current sample? 

  9. My hypothesis is that St. Catherine students are more anxious than depressed. Is it true based on my sample?

• Processing speed of children were tested in two tasks. Their first task was simple: they had to decide if the pictures presented were animals or not. Their responses in seconds were: 7; 8; 6,5; 3,5 and 5. The other task was more difficult. Here they had to answer ten questions as fast as they could. Their responses in minutes were: 4; 5; 3; 5; and 2. Is there a significant difference in their performance in the two tasks? 

Topic7: Correlation and regression

1. Open the StatisticsPracticalTopic07_Hourssutdiing database

2. CORRELATION 

   1. Is there a linear relationship (=correlation) between how many hours one studies and the grade based on the current database?
      r = 0.778; p = 0.023

Based on the Pearson correlation I have run, r = 0.778, p = 0.023, there is a significant, strong, positive linear relationship between how many hours one studies and their grade. This results suggest, that those who study more for the exam, get better grades. 

2. 2. there are 3 important parameters of correlation analysis: significance, strength, direction 

Correlation Coefficient (r) - same ranges for the positive values: 

-0.2 to – 0.4      Weak - association

-0.4 to -0.6        Moderate - association

-0.6 to -0.8        Strong - association

-0.8 to -1.0        Very strong - association

2. 3. Practice correlation analysis: Is there a linear relationship between the student's satisfaction with their grades and their grade? Hypothesis: there will be a strong positive correlation. Those who got better grades will be more satisfied with their grades. 

   4. Practice correlation analysis: Is there a linear relationship between the student's expected grades and their actual grades? Hypothesis: there will be a strong positive correlation. Those who expected better grades got actually better grades. 

3. GRAPH the relationship: 

   2. 1. Graphs – Scatter/Dot – Simple

      2. Line of best fit: Fit line at total

4. Regression analysis - for predicting 

   1. Predict the average if 10 hours are spent with of studying/week. - see ppt

5. Practice regression analysis - for predicting

   1. If Elizabeth expects a grade 5, how many hours should she spend with studying? 

1. open the StatPracticalChiSquare database

2. Is there a significant linear relationship between novelty seeking and risk taking on the current sample? 

3. If somebody would score 11 on the novelty seeking scale, what would be his score on the risk taking scale? 

4. Chi square tests: 

   1. Is problematic online gaming significantly more frequent amongst males, as compared to females on the current sample?
      Based on the Chi-square analysis on the current sample, problematic online gaming is not statistically significant amongst males as compared to females. Chi square (1)= 2.967 p = 0.085. Also report the frequencies in percentages!! Out of the 8 male participants, 12.5% is non-problematic online gamer, 87.5% is problematic. Out of the 12 female participants, 50% is non-prolematic online gamer, while 50% is problematic online gamer. 

   2. Is it true that problematic online gamers are also more frequently regular smokers? - answer the question based on the current sample

   3. Are females more likely to smoke regularly? - answer the question based on the current sample. 

   4. Is participating in extreme sports more frequent amongst problematic online gamers? 

5. Is the mean novelty seeking score significantly higher in the regular smokers as compared to the non-smokers? 

1. open the StatPractice_UniversityStudents database

2. Is there a linear relationship between the student's current satisfaction level and their general life satisfaction? 

To test the relationship between the student's current satisfaction level and their general life satisfaction, we run a linear correlation test. We found a positive, very strong correlation, and it crossed the significance level (r=0,936, p<0,001). 

3. Is it true that Europeans more frequently have roommates than non-Europeans? 

Based on the Chi square test, we found a non-significant value (Chi-square (1) = 1,222, p=0,269).

Out of my 13 European participants, 61.5% has roommates, while from my 17 non-European participants, 41.2% has roommates. We can conclude that Europeans are more likely to have roommates, however, the ratio differences are not significant

4. Is it true that students living on campus are more likely to have roommates than off-campus living students? 

5. If a student would score 10 on their current level of satisfaction scale, what would be his score on the general satisfaction scale? 

To predict the general satisfaction scale of a student that score 10 on the current satisfaction scale, we ran a regression test and found that they would score 10,7.

6. Is it true that current life satisfaction negatively correlates with the homesick scale score? Are those students who are more satisfied with their current life less likely to be homesick? 

7. Is there a significant difference between the European vs non-Europeans in the frequency of smoking? Is smoking more frequent amongst non-European students? 

8. What would be somebodies score on the homesick scale, if he would score 3 on the general satisfaction scale? 

9. Is it true that BA students are more likely to have roommates than Ma students? 

To check the frequencies of having roommates amongst BA and MA students, we run a Chi -square test (Chi-square(1) = 2,143, p=0,143), and it shows no significance amongst the ratios. Out of my 14 BA students, 64.3% of them has roommates, while out of 16 MA students, 37.5% of them has roommates. We can conclude that is more likely for BA students to have roommates, however, these ratios are not significant 

10. Is it true that non-Europeans are significantly more likely to come to Hungary to study on an MA level than an o BA level? 

We found a strong, positive and significant correlation between the general satisfaction scale and the homesick scale. To check that, we ran a linear correlation analysis (r=,866, p<0,001).

11. s there a linear relationship between the student's general satisfaction level and their homesickness scale score? 

12. Is it true that BA students are more likely to live on-campus than MA students?