Child psychologist: effect of violent vs non-violent game on children's later violent behaviour.

• 2 experimental conditions (violent / non-violent). DV = subsequent violent behaviour. Design = independent measures.
• If repeated measures: order effect — aggression from the violent game could linger into the non-violent condition ⇒ use counterbalancing.
• Independent measures risk participant variables (original level of aggression) ⇒ random allocation or matched pairs (match on aggression).
• Avoid demand characteristics: children shouldn't know it is an experiment — e.g. say it is a computer-game competition.

Extra level with non-violent pictures = a control condition? Playing no game would be control; pictures = another experimental level. Bored of games by 2nd condition = fatigue effect.

Does watching pro-social TV make children kinder? DV = times child helps a crying doll in 10 min of video.

• Lab: each child sees helpful-characters cartoon or neutral cartoon, then given a doll.
• Field: parents show cartoon at home; child videoed with doll.

Field = more ecological validity (familiar home, less nervous) ⇒ results generalise to school/park. ✗ But distractions at home may stop children paying attention to the cartoon.

• IV = age group (older / younger adults); DV = how quickly they press a button after recognising the emotion on a face.
• Standardisation: same distance from screen for everyone.
• Pilot study: timing by stopwatch was inconsistent ⇒ switched to computer ⇒ ↑ reliability.
• Validity: faces of both younger & older people used. Controls + standardisation ⇒ colleague can replicate exactly.

Aim: do hard or soft chairs help students work better?

• Non-directional: There is a difference in work rate of students on comfortable vs uncomfortable chairs.
• Directional (1): Comfortable chairs ⇒ higher work rate (comfort → happier → work harder).
• Directional (2): Comfortable chairs ⇒ lower work rate (too comfy → sleepy/lazy).
• Null: Any difference in work rate on comfortable vs uncomfortable chairs is due to chance.

• Experiment: lab; IV = type of recording/condition; DV = pulse rate (should fall); design = independent measures (allocated to one level).
• Consent asked only for part of the procedure (imagery-based relaxation recording; pulse monitors & dark rooms) ⇒ deception (false info: "some people see disturbing flashing lights").
• Right to withdraw given at the start. Privacy: told they won't be asked about intimate thoughts. Confidentiality: ID numbers instead of names.
• Not protected from harm: intimate imagery or flashing-lights false claim could distress. Might withdraw for the same reason.
• Deceive to avoid demand characteristics (they'd alter their pulse if they knew the aim).
• Bad hypothesis "imagery-based recording will help people relax" — variables not operationalised (what recording? how is "relax" measured?) and no comparison condition.

Online questionnaire students did in free time. Closed Qs 1–4 (quantitative): homework always/sometimes/never; "written own glossary?" yes/no; "'Psychology is a difficult subject'" Likert 5-point; rate 0–6 how well you understand. Open Qs 5–6 (qualitative): what do you do after each lesson to remember? how will you plan revision?

• Colleague helped interpret open answers; a coding list was devised ⇒ inter-rater reliability (e.g. Q5: reading/copying notes, textbook, online, friends, teacher; Q6: copying notes, summary notes, mind maps, past papers, test cards).
• Why online? students could do it in free time ⇒ saves lesson time, no need to collect/mark paper, may be more honest/anonymous.
• Open vs closed: closed gives numbers; open gives reasons.
• Take steps to raise inter-rater reliability because different people would otherwise interpret the same answers differently ⇒ findings inconsistent/subjective.

• Plan 1 = structured interview (specific closed Qs: "how many times has someone held a door? once/twice/3+", "Have you helped carry shopping? yes/no") — he wants objective, countable data; data from closed Qs needs no interpretation.
• Plan 2 = unstructured: begin "describe how friendly/helpful people are…", then follow up. More depth but subjective — he already believes the layout increases helpfulness ⇒ may interpret in a biased way.
• Dr McArthur's suggestion: semi-structured — fixed comparable questions + room to expand.

• Closed: "Which animals scare you? dog / spider / cat…" · Open: "Describe how you feel when you see a spider."
• Why unstructured? follow up on individual answers, explore reasons for a person's phobia in depth ⇒ more valid.
• Ethical problem: talking about phobia may cause distress ⇒ protection from harm / may feel unable to refuse a personal question ⇒ privacy; remind of right to withdraw / ignore questions.

• Methods/techniques: video in sleep lab (sleep-talking/walking) · EEG (brain waves; detect dreaming) · wake & ask emotional state/dream · same test when not dreaming · dream diary (self-report) · interviews with family (when/how often) → ≈ 5.
• Why a case study: one patient (single instance) + detailed, in-depth data using several techniques.
• Withdrawal difficult as researcher has built a close relationship; lots of invested time/records; PMA may want treatment/help from the clinic.
• Ethical reason for "PMA": confidentiality (not privacy!) — identity hidden, using initials.
• Other ethical issue: privacy (filming/sleeping/personal dream content) or protection from harm (woken repeatedly, distressing dreams).
• Methodological issue: low generalisability (one person) or subjectivity (researcher involved).

• a) three techniques: interview, observation (of how they respond), questionnaire (to family/friends), records/tests.
• b) ethical problems: confidentiality (detail makes identity easy to work out); informed consent (may be hard if limited understanding / vulnerable); privacy (personal questions); protection from harm (distress when recalling). Solution (one): anonymise with initials/change details; check understanding & use presumptive/guardian consent; allow right to ignore Qs.
• c) No — a single individual, so generalisability is limited; may not typify others.

• Controlled observation: three students act as confederates, take turns to look sad in the common room at lunchtime.
• Dr Huang + colleague on lunch duty = non-participant, overt observers (students used to them ⇒ ignore them).
• Structured with a list of behaviours — improves inter-observer reliability (same definitions):
   • shows concern without action — looks at confederate, doesn't move towards them
   • verbalises concern — goes up & asks "are you OK?"
   • takes action — e.g. arm round them / buys a drink.

Activity 1.9 ideas: extra category e.g. "ignores" (walks past without looking). Unstructured version = record everything students do. Naturalistic alternative: just watch for anyone who is really sad — more ethical: no acting/deception; less ethical: filming/watching private distress without consent. Covert participant: sit among them disguised as a student chatting (observer's role hidden).

• a) Debra = naturalistic (animals in own environment); Jin = controlled (artificial situation, objects in a special box).
• b) Debra hides in a tree ⇒ animals unaware ⇒ covert (non-participant).
• c) e.g. Jin: structured — rats' interactions with objects can be coded into categories (sniff / touch / pick up…) ⇒ reliable. Debra could use unstructured first to find out what animals do, then structure.

Co-variables: years phobia has lasted and severity rated 1–10 (how much it interferes). Positive correlation = longer phobias more severe. No relationship ⇒ scatter points randomly placed, not on a line.

• Negative alternative = longer-lasting phobias are less severe (as years ↑, severity ↓).
• Complete hypothesis: There will be a negative correlation between the number of years a person has had a phobia and the severity rating (1–10) of the phobia — operationalise both.
• Positive r found ⇒ can't say time makes phobias worse (correlation ≠ causation; e.g. more severe phobias may last longer, or a third factor such as anxiety level).

• a) Correlational: two measured variables; neither is manipulated by the researcher (no IV/DV, no conditions).
• b) Either; e.g. negative — more coffee ⇒ poorer/lighter sleep ⇒ fewer dreams recalled (justify whichever you choose).
• c) measure coffee: ask participants to keep a diary of cups per day (or questionnaire). Weakness: self-report may be inaccurate (forget, cup sizes differ, social desirability); strength: easy / quantitative.

• a) contact at sensible long intervals e.g. every 1–2 years (career progress changes slowly; too often ⇒ boredom/attrition).
• b) happiness in career: rating scale / Likert-type questionnaire (1–10) (quantitative) or interview with open Qs (qualitative).
• c) collect contact details (phone/email/address, a second contact person) at the start & keep updated; maybe incentives/newsletters; stored securely (confidentiality).

• IV "hard/soft chairs" (or "comfortable") still vague ⇒ "chairs with wooden/plastic seats vs chairs with padded seats".
• DV "work better/harder" vague ⇒ expand "work rate": number of pieces of homework handed in late or time spent doing extra work.
• Confounding variables: e.g. different room lighting (situational variable — aspect of environment); hard-working students (participant variable); sunny-day weather (situational); if students choose chairs, lazy ones pick comfy ⇒ uncontrolled variable making soft chairs look worse. Subject groups: all soft-chair class arts students, hard-chair class maths students ⇒ if DV is maths test, hard-chair group wins because of ability (participant variable), not chairs.
• Activity 1.11 ideas: IV = "sitting on a cushion vs not"; DV = number of questions completed in 20 min; uncontrolled variable = time of day / noise / type of lesson.

• Amount of sleep = hours & minutes spent unreactive, or identified by EEG (REM / non-REM).
• Emotional reactivity = how likely a person is to respond negatively to innocent comments / how often they argue (measured by questionnaire).
• Alternative for "emotions" (Activity 1.11): rate mood on a 1–10 scale, or count the number of times they cry/laugh in a day.

• 10a IV = revising alone vs with other people (setting); DV = how effective revision is, e.g. score on a test after revising.
• 10b Directional (one-tailed) — predicts which condition is better (alone).
• 11a(i) Others motivating ⇒ work longer ⇒ better test score in group ⇒ hides/reverses the effect of "alone" (systematic confounding). (ii) control: fix revision time (all revise 30 min).
• 11b(i) Distraction ⇒ group works less effectively ⇒ exaggerates alone advantage. (ii) Control: same quiet room, set task, ban phones/talking off-topic.
• 11c Student ability = random uncontrolled variable (spread across both groups) ⇒ less important than systematic confounders (which affect one level only); random allocation helps.

• Opportunity (your classes) ⇒ all science students ⇒ may be especially good at logic tasks ⇒ not representative of college.
• Volunteer (canteen announcement) ⇒ may not give a range of subjects.
• Random: exam candidate numbers → random-number generator → first 30 numbers.

Activity 1.12: neighbours (mainly retired) ⇒ opportunity ⇒ biased to old age · younger students in detention ⇒ volunteer sample missing them ⇒ biased · Professor Chao social media = volunteer · Dr Singh asking all residents of two care homes = opportunity · Dr Huang numbers from hat = random.

• a) Opportunity (Kiaria): only friends/available students ⇒ similar ⇒ not representative of all students' activities.
• b) Volunteer (Haniya): only those who like/are already active in non-academic things (and have time) respond ⇒ biased.
• c) Random (Habib): need a complete list of all students; by chance may pick few who do non-academic work / mainly one type; hard to contact everyone selected.

• Dr Singh: quantitative — measure pulse rate + rating scale (very / quite / not at all stressed).
• Prof Chao: qualitative — interview; Qs: "How do you feel if you think someone is standing too close?" "Describe how you felt the last time a professional (doctor/dentist/optician) was very close."
• Her Qs are open. More ethical extra Qs: open — "Describe what you do if someone stands too close"(qualitative); closed — "Do you mind being sat next to strangers? yes/no" (quantitative). (Avoid distressing/personal content.)

• Subjects: Maths 4, English 6, Psychology 10 ⇒ mode = Psychology.
• Homework night — boys: Sunday has 11 (highest) ⇒ boys' mode = Sunday. Girls: Friday 6 & Sunday 6 ⇒ two modes. Combine totals: Sunday 11 + 6 = 17 ⇒ overall mode Sunday.
• Facial expression test: "which face looks happiest?" — face chosen most often by each group = mode.

AS (ranked): 1,2,2,3,3,3,5,5,5,6,6,6,6,7,8,8,8,9,9,10  ·  A Level (ranked): 4,6,6,7,7,7,7,7,7,7,9,9,9,9,9,10,10,10,10,10

A Level median (8) & mean (8) > AS (6 & 5.6) ⇒ A Level students think they work harder. Modes: AS = 6, A Level = 7.

SD: AS = 2.62, A Level = 1.72 ⇒ AS students' opinions are more varied (most A Level say "working hard"; AS opinion ranges from "not hard" to "very hard").

• Range's problem: doesn't reflect outliers. If the lowest A Level rating (4) became 1: mean 7.85 (not 8) but range 7 → 10 = same as AS, though within groups AS still more varied.
• Standard deviation (s, SD, σ) = average difference of each score from the mean (the deviation); bigger SD = more varied. It uses every score, so isn't distorted by outliers like range: SD A Level 1.72 → 2.16 with that change — still below AS 2.62.

Jamal: 0,0,3,4,5,6,7,8,9,10 · Nurul: 0,1,1,2,2,2,3,4,4,5

• Mode misleading (0 → "horrified"); median 5.5 feels better; mean 5.2 is lower because the two zeros count with their absolute value but not for median (just "lower than central values").
• Median & mean ⇒ Jamal's group did better; but Jamal's modal (0) < Nurul's (2) ⇒ mode alone suggests opposite!
• Range 11 vs 6 ⇒ Jamal's results more spread. SD 3.49 vs 1.58 ⇒ Jamal's group more varied.

Activity 1.14 Q1: Face A 0, B 3, C 7, D 1 ⇒ modal response = Face C.

3, 45, 32, 14, 21, 18, 38, 29, 14, 39

e) Histogram axes: x = "Number of words said at 18 months (grouped 0–10, 11–20, 21–30, 31–40, 41–50)" with title; y = "Number (frequency) of infants" (0–4).

• A: averages of "thank you" (boys 6, girls 7) ⇒ bar chart (categories boys/girls).
• B: helped in situations (pram 14, elderly 7, child 16, injured person 3) ⇒ bar chart (discrete categories).
• C: % students responding in 0–5, 6–10, 11–15, 16–20, 21–25 s (0%, 15%, 45%, 20%, 5%) ⇒ histogram (continuous seconds).
• D: "times lent notes" (5,3,9,4,2,6,9,7,3,4) vs "likelihood to give lunch" (7,2,7,10,1,8,8,7,4,5) ⇒ scatter graph (two co-variables ⇒ correlation).

• Q13 table for Fig 1.18 (title: "Time spent tidying per day by age group"): columns = Age group (years) 0–5, 6–10, 11–15, 16–20, 21–25, 26–30, 31–35 · row = Minutes per day tidying (read each bar height off the graph; empty groups = 0). Units go once in the heading.
• Q14 bar chart of Table 1.5: x = student group (A Level, degree level); y = average time working per week (hours); bars 6 and 14; title e.g. "Average time spent working per week by A Level and degree students".
• Q16 Ethelred: negative correlation ⇒ scatter points from top-left to bottom-right: x = hours of sleep, y = happiness; line of best fit sloping down.

• Aim · IV with operational definition of each level · DV operationalised + how measured · experimental design + justification · controls + why · sampling technique, size, special features · materials/stimulus lists · questions (closed/open + scoring) · ethical issues + steps · raw-data table · descriptive-stats table (include central tendency) · graph axes (title, labels, units).
• Do NOT conduct it without explicit teacher approval.

• Replacement: considered videos of wild animals but rejected — can't be sure what they react to ⇒ findings not valid.
• Species: rats — bred for lab use, can be housed alone without distress.
• Number: repeated measures design ⇒ fewer animals than independent measures.
• Pain/distress: food they liked (peanuts, "approach") vs disliked (lettuce, "avoid"); not deprived of food (fresh food in cage before testing).
• Activity 1.16: aversive stimuli (loud noises, electric shocks) unethical — cause pain & distress. Repeated measures problem: order effects (e.g. rats learn the task / tire) ⇒ counterbalance.

• 17a Rashid's study: deception (pretended help needed), no informed consent, people may be distressed/embarrassed/risk near the road ⇒ protection from harm.
• 17b Debrief afterwards, keep observations confidential/anonymous, make sure the situation is safe, use presumptive consent from similar people.
• 18a Dr Hadid should: use the minimum number of mice, house them with enough space/food/water (social species ⇒ in groups), use preferred food as reward not deprivation, pilot study, seek vet advice.
• 18b should not: deprive of food/water, use aversive stimuli (shocks), isolate the mice, or use more animals than needed.

• Field experiment + structured observation; same male warden in both conditions (control ⇒ valid); 4 observers, one per direction.
• Operational definitions (⇒ reliable): Obedient: slowing down (visibly reducing speed) · stopping (halt before line). Disobedient: stopping late (halt past line) · driving on (fails to stop when told).
• Activity 1.17: generalisability — repeat in a village as well as town (different place) and with a female warden (different gender) ⇒ findings apply more widely. High ecological validity because drivers are in a real, normal setting and unaware (no demand characteristics). "Visibly reducing speed" is subjective — what is "visible" differs between observers. Measure inter-observer reliability because if observers disagree, data are inconsistent/invalid.
• Ethical problem: debriefed drivers find out they were studied ⇒ may feel deceived/embarrassed/distressed; consent not obtained.
• Giving a number to call (instead of taking theirs) = right to withdraw (they choose whether to take part further and can withdraw their data) — they weren't pressured.
• Reasons for stopping given by phone: two colleagues interpret on a numerical scale of how strongly the driver felt punishable ⇒ strong positive correlation of their scores ⇒ high inter-rater reliability (they interpret consistently).
• To replicate need: junction location/type, time of day & traffic, exact clothing, wording of warden's instruction, line position, how observers recorded.

• a) Double-marking with a detailed mark scheme ⇒ improves reliability (inter-rater: two markers consistent; objective). It also helps validity only indirectly.
• b) Students sharing tests = confounding variable / contamination ⇒ threat to validity (less-frequent group got the IV partly) — IV not truly separated.
• c) Improve generalisability: use different classes/schools and ability levels, bigger and more varied (random) sample, different subjects.

• a) Alternative hypothesis — it predicts a difference in the DV (mindfulness) between conditions (training vs waiting), not that it is due to chance. [1]
• b) Directional (one-tailed) — says which condition will have greater mindfulness (training). [1]

• a) Strength (questionnaire): easy to analyse — closed Qs give totals/averages ⇒ summarise. Reason/elaboration: can be given to many people ⇒ larger sample. Weakness: social desirability bias — may answer to look acceptable ⇒ ↓ validity; or low return rate / open Qs need interpretation. [4]
• b) Closed: "Do you think people with phobias should be helped? yes / no". Open: "Explain why you think it is / is not important to help people with phobias." [2]
• c) Mainly reliability — interpretation of open answers may be inconsistent between researchers/occasions (inter-rater reliability), so results not the same each time. [2]

a) Describe how she could conduct it — a model plan (mention every item):

• Type & design: field experiment in school or lab; independent measures (IV = gender: boys / girls) — can't allocate, so natural groups. Aim: boys vs girls doodle more.
• DV operationalised: number of doodles (or time/area doodling) during a 10-min boring recorded talk/task; count with a tally.
• Hypothesis (e.g. two-tailed): there will be a difference in the number of doodles by boys and girls.
• Sample: e.g. 20 boys + 20 girls from her school; random sampling with numbers from a hat (or opportunity/volunteer).
• Controls & standardisation: same task, room, time of day, length, instructions, same paper & pens; no talking.
• Ethics: informed consent (parents too if under 18), right to withdraw, confidentiality (numbers not names), debrief.
• Collect data → mean doodles per gender → compare (bar chart).

b) Weakness + improvement: e.g. participants may notice the aim (demand characteristics) → say it is a study of listening and measure doodling unobtrusively; or participant variables (artistic skill) between groups → matched pairs on artistic ability / larger random sample. [4]

• a) IV = type of animal (parrots vs cats). [1] b) DV = time (seconds) taken to reach the food inside the box. [1]
• c) Opportunity sampling — he is using the 3 parrots and 2 cats he happens to have/is available (not selected randomly from all animals). [2]
• d) Mean (uses all values) — or median if outliers. [1]
• e) Validity: time may depend on sense of smell, not intelligence — the test isn't measuring what it claims (confounding variable). [2]

• a) Volunteer (self-selected) sampling — participants responded to adverts and chose themselves. [2]
• b) Generalisability: sample small (10), 9 females 1 male, mainly retired ⇒ unrepresentative; findings can't be applied to men, young people, etc. [2]
• c) Deception — they are told it will help them sleep / keep them awake (false info); may affect consent (or distress, right to withdraw). [2]
• d) Minutes asleep is more reliable — objective, measured with a clock/same unit each time; self-report is subjective/may be inaccurate & influenced by what they were told. [2]